Northwind Health · Executive Leadership Team

91 people use Claude. 3 ship code.
The shape of Northwind Health's AI deployment, and who should lead Sales and CS next.

Companion read to AI Transformation at Northwind Health ([CEO] / Crawford / [Tiger Team Builder])
Data window: 2026-01-05 → 2026-05-06 · 95 active · 4,173 conversations · 1,040 commits
Prepared by Blueprint · May 17, 2026

01 · Where we are

Joss's deck reports the outcomes. This one reports the shape of the work underneath them.

Marketing
seven-figure
Marketing — AI-sourced ARR pipeline
19 opps · Q2 through May 6 · Auto Intent $755K
Niven · Micah · Lane
Compliance
~5×
Compliance — AI efficiency multiplier
thousands of AI hrs vs 930 worked hrs
Hollis
Learning
several hundred
Learning — hours saved per month
writing · updates · metadata · translation
Sterling
Sales
mid-six-figure/yr
Sales — projected productivity (centralized)
30 reps × 5 hrs reclaimed
Jamie
Read this first

Three people ship production code with Claude. Eighty-eight more chat with Claude every week and ship nothing. That is your hybrid model in one sentence: a small Tiger Team builds, a long tail consumes. The four KPIs above are real, and they are produced by a much smaller group than the deck's adoption story suggests. The named federated leads — Sterling, Hollis, Niven, Micah, Lane — produce their results in completely different shapes of usage. Sterling saved 330 hours per month with 57 chat messages. Micah shipped low-hundreds of commits with 32. The next Sales lead and the next CS lead need to fit one of those shapes. Hiring against generic "AI fluency" will pick the wrong person.

02 · Adoption curve

Daily Claude.ai conversations across the company

2026-012026-022026-032026-042026-05
Top 5 contributors by message volume
Blair
3,624
Quinn
3,510
Eden
2,626
Nolan
2,307
Phoenix
2,086
Three people produce most of the depth. The line goes up when a Tiger Team skill ships, then flattens. The flat stretches are the rest of the company waiting for the next thing to use. That waiting is what the Functional AI Lead role is supposed to absorb — the lead's job is to keep their department moving when the Tiger Team is busy.
03 · Anthropic API usage and infrastructure

$2,564 over four months for an infrastructure that produced the seven-figure Marketing pipeline and 5,090 churn dossiers. ~$513/month total.

This section is the routine usage report on the Anthropic stack: what got spent, where, and the cost + security discipline now in code or pending your approval. The numbers are small; the disciplines below are what keep them that way as the agent stack scales.

2026-012026-022026-032026-042026-05 $671 · 2026-03-26 churn analysis run
The shape of the bill
Total — Jan 5 to May 13
$2,564
Average month
$513
Median day
$6
Highest day — March 26, the 5,090-account churn analysis
$671

Most days fall under $20. March 26 produced the account-dossier corpus that powers Sections 04, 05, 07 and 09 of this report — the churn analysis that processed every active and cancelled account through an Opus agent. Cost discipline in the orchestrator (caching, defaults, pre-flight) keeps similar future runs an order of magnitude smaller without losing the depth.

Cost discipline — in code
  • Pre-flight cost estimator on every batch submitter. Eight scripts compute the dollar estimate before submitting — the agent swarm plus seven campaign batch jobs. Helper at tools/llm-cost-guard/: prints summary at any spend, requires interactive confirm at $100+, requires a typed dollar amount at $1K+, requires a second flag at $10K+. Any new orchestrator written into either repo inherits this gate.
  • Prompt caching on shared system prompts. cache_control: ephemeral wired into the agent-swarm orchestrator's system block. First call in a batch pays full rate; subsequent calls pay 10%. On the next run of the same churn-analysis workload, this alone cuts the bill ~80%.
  • Sonnet 4.6 as the default everywhere. Opus is opt-in by explicit flag and automatically triggers the cost check. Sonnet is 5× cheaper for input, 5× cheaper for output, and matches Opus on most agent tasks the team runs.
  • Hard rules in CLAUDE.md at three levels — global (~/.claude), Blueprint repo, the-team-repo repo. Any orchestrator code that skips the pre-flight or skips caching gets caught at review.
Security & key hygiene — needs your sign-off
  • Move 8 of 9 API keys into workspaces with daily caps. Today every key but one runs against the org default budget with no ceiling. Recommend a Tiger-Team-Production workspace at $5K/day for production keys, Agent-Swarm-Experimental at $500/day for one-offs. Caps fire automatically when crossed. Console → Workspaces.
  • Revoke 7 dormant API keys. Of the 9 named keys, 7 have produced effectively zero traffic in this window. Each is a live credential someone could find in an old config file or chat log. Console → API keys → rotate or delete.
  • Set budget alerts at $1K/day and $5K/$10K/$25K monthly. Wired to a Slack channel via the org webhook. An unexpected day surfaces within minutes, not at month-end. Console → Billing → Alerts.
  • Rotate any key that's been pasted into a chat, doc, or screen-share. Treat the admin key as the highest-sensitivity credential — it can move money. Quarterly rotation as routine; immediate rotation on any visibility leak.
  • Enable the Claude Code OpenTelemetry exporter. The per-engineer Claude Code analytics endpoint returns zero records today, so the org has no native visibility into which engineers ship with Claude Code. Configure once at the org level; data flows from then on. Console → Settings → Analytics.
Forward operating practices

Per-session budget caps in user-facing agent UIs. Sales Owl, the BDR dashboard, and any customer-facing agent need a per-user-session ceiling — refuse the (N+1)th Opus call. Route first-pass to Sonnet or Haiku; escalate to Opus only when the upstream agent explicitly asks. Same discipline as the orchestrators, enforced at the application layer.

Quarterly Anthropic-stack audit. Re-pull the admin API data and review four things: model mix (Opus should stay a minority of spend), cache_read ratio (healthy agents land 50–90%), dormant-key count (revoke if >30 days idle), single-key concentration (one key driving 99% is a bus-factor risk). Script lives at scripts/00_pull_admin_api.py in this skill.

Decide the org-seat policy. Today only Riley is on the Northwind Health Anthropic org. The 91 chat-active employees in this report run on personal Claude accounts and never show up in admin analytics. Inviting them costs per seat but gives one auditable ledger and unlocks per-user Claude Code analytics. Worth a 30-minute conversation with Riley on the trade-off.

04 · The shipping gap

Of 98 people who use Claude, only 3 ship production code.

The chat-only 88 are the bench for every future Functional AI Lead. They know the workflows. What they lack is the data infrastructure Tiger Team is meant to build for them. The pace at which Tiger Team unblocks them is the speed of the federated model.

PersonCommitsBy repoLinesChat msgsActive days
River
person-034@northwindhealth.example
461 CO 435 · MT 26 +9,496,270 / −121,075 1,144 62
Micah
person-024@northwindhealth.example
235 CO 235 · MT 0 +154,816 / −6,530 423 22
Lane
person-020@northwindhealth.example
121 CO 64 · MT 57 +1,229,778 / −13,931 32 5
Jordan CrawfordBlueprint partner
SantaJordan@.noreply.github.com
173 CO 77 · MT 96 +26,206,067 / −4,082,969
Chat-only top six. Finn (3,624 msgs, 60 active days), Quinn (3,510 / 45), Wyatt (2,626 / 54), Rowan (2,307 / 58), Cameron (2,086 / 22), Elliot (2,059 / 28). These six together produced more chat volume than the three shippers combined and have zero commits to show for it. Two questions to ask each manager: (1) is their AI work supposed to ship, or is it analysis that informs a human decision? (2) if it's supposed to ship, what is the Tiger Team owed by when?
About the two repos. The team started in the-legacy-repo (January through mid-March) and then cut over to the-team-repo (late March onward). Both are counted in every commit number above.
05 · The leaderboard

All 91 active, ranked against Joss's six attributes

Each user was scored 1–5 by Claude on Joss's six slide-14 attributes — Deep Expertise, Creative Problem Solver, Organized, Driver, Trusted Peer, Learns Fast — using their actual conversations and commits as evidence. Two reads matter: the composite score (rank) and the radar shape (where their strength lives). ● marks the people who already ship production code. The shape on the radar is what predicts whether a person can be a Functional AI Lead. A spike on Driver alone is a lone wolf; a balanced shape with strength on Trusted Peer is a coach.

#PersonActivityRubricScoreRead
1
River
person-034@northwindhealth.example
Engineering Builder
1,144 msgs · 149 convs · 62d
commits 461 · +9,496,270/−121,075
4.8 Riley is Northwind Health's most prolific AI builder by a wide margin — hundreds of commits, 8 projects, and a monorepo spanning Gong/Salesforce/Outreach MCP servers, AI CRM queues, and executive dashboards, all self-directed in under five months. He combines deep RevOps and go-to-market data fluency with genuine engineering output, making him the clearest candidate for a formal Functional AI Lead role.
2
Micah
person-024@northwindhealth.example
Marketing Builder
423 msgs · 32 convs · 22d
commits 235 · +154,816/−6,530
4.5 Micah is Northwind Health's most technically accomplished sales operator — he has shipped a production BDR dashboard (low-hundreds of commits, +100k lines) that integrates Salesforce, intent data, Gong, and AI-personalized outreach at scale, all driven by deep domain knowledge of the BDR workflow. He operates with full builder autonomy, learns new tooling rapidly, and is already the kind of Functional AI Lead the transformation program is designed to find.
3
Lane
person-020@northwindhealth.example
Marketing Builder
32 msgs · 6 convs · 5d
commits 121 · +1,229,778/−13,931
4.3 Lane is Northwind Health's most prolific AI-native builder, shipping around 120 commits including a full multi-agent tiger-team orchestration pipeline in under three weeks with minimal oversight. He pairs strong GTM domain intuition (hospital EMR strategy, CMP reinvestment positioning) with hands-on engineering output, making him one of the highest-leverage AI leads in the company.
4
Robin
person-035@northwindhealth.example
Sales Power User
1,219 msgs · 73 convs · 31d
commits 0 · +0/−0
3.8 Riley is Northwind Health's most prolific AI user in the window, running around 1,000+ messages and a couple thousand tool calls across 50 days to produce polished, deal-specific deliverables — pricing guides, competitive business cases, negotiation decks, and territory audits — entirely self-directed. His deep command of prospect pain points and competitor mechanics elevates his AI output well above generic; the main gap is formalizing his work into shared templates others can inherit.
5
Eden
person-061@northwindhealth.example
RevOps Workflow Embedder
2,626 msgs · 283 convs · 54d
commits 0 · +0/−0
3.7 Eden is Northwind Health's most operationally fluent AI user, embedding Claude directly into RevOps workflows — Salesforce architecture, comp plan design, deal desk process, and CRO-level reporting — with impressive depth and consistency across 54 active days. She is a clear Functional AI Lead candidate for RevOps; the next step is formalizing her outputs as shared, maintained artifacts and spreading her methods to peers.
6
Tobin
person-077@northwindhealth.example
Compliance Workflow Embedder
1,020 msgs · 75 convs · 37d
commits 0 · +0/−0
3.7 Tobin is Northwind Health's most prolific compliance policy builder, running sustained AI-assisted audits across ASC, FQHC, SNF, and HHA settings with demonstrable mastery of federal and accreditation standards. His prompt sophistication is rising sharply, but impact would compound significantly if his outputs were committed to shared systems and made visible to peers.
7
Logan
person-022@northwindhealth.example
Sales Power User
778 msgs · 124 convs · 50d
commits 0 · +0/−0
3.7 Logan is Northwind Health's most prolific sales AI user, consistently turning prospect intelligence and pricing data into polished executive deliverables that would otherwise take hours to produce manually. He's also emerging as an internal AI advocate, having already pushed Claude access to peers — making him a natural candidate for a formal Functional AI Lead role on the Sales team.
8
Hollis
person-014@northwindhealth.example
Compliance Workflow Embedder
602 msgs · 54 convs · 31d
commits 0 · +0/−0
3.7 Hollis is Northwind Health's most domain-dense AI user in compliance, consistently converting complex regulatory frameworks — HRSA OSV, HazCom 2024, CMS AI Playbook — into structured, deliverable-ready artifacts with minimal iteration. He operates as a self-directed practitioner whose output quality is exemplary, though peer influence and system-level execution (commits, shared templates) remain the next growth frontier.
9
Blair
person-003@northwindhealth.example
Sales Power User
3,624 msgs · 293 convs · 60d
commits 0 · +0/−0
3.5 Blair is Northwind Health's most prolific AI-powered seller, using Claude daily to run end-to-end deal cycles—from cold call script factories to live pricing negotiations and closed-lost recovery—with a depth and volume that is genuinely exceptional. His impact is currently personal rather than organizational; unlocking his potential as a peer coach or template author could multiply his influence across the sales floor.
10
Quinn
person-031@northwindhealth.example
Sales Workflow Embedder
3,510 msgs · 220 convs · 45d
commits 0 · +0/−0
3.5 Quinn is a high-volume Sales manager who has deeply embedded AI into daily BDR operations — building prospecting lists, re-engagement trackers, and performance reviews at scale across his entire team. His functional Salesforce knowledge and coaching instincts are strong, but formalizing his output into shared projects and documented systems would unlock significantly more organizational impact.
11
Lennox
person-021@northwindhealth.example
Sales Power User
1,597 msgs · 76 convs · 29d
commits 0 · +0/−0
3.5 Lane is Northwind Health's most prolific individual AI consumer, having built a near-daily AI-assisted sales workflow that spans Salesforce, Gong, and email to manage his full RSM pipeline with remarkable depth and consistency. His impact is currently self-contained — the next unlock is channeling this fluency into shared templates or peer coaching to multiply his influence across the sales org.
12
Frankie
person-012@northwindhealth.example
Sales Power User
994 msgs · 53 convs · 22d
commits 0 · +0/−0
3.5 Frankie is Northwind Health's most prolific AI-powered sales contributor, using Claude daily to turn raw prospect research and call transcripts into polished executive summaries, RFP responses, and competitive battle cards. Her domain expertise in healthcare LMS competitive dynamics is exceptional, and she's the clearest example of AI deeply embedded in a sales workflow — the missing step is externalizing that into shared assets her peers can replicate.
13
Oakley
person-028@northwindhealth.example
Compliance Workflow Embedder
931 msgs · 27 convs · 15d
commits 0 · +0/−0
3.5 Oakley is a compliance specialist who has embedded AI deeply into her regulatory policy workflow, demonstrating rare command of CFR citations, California Title 22, and bilingual FQHC standards across nearly 1,000 messages in two months. She operates with strong methodical discipline and self-imposed quality frameworks, though her impact remains largely individual with no visible project authorship or cross-team influence yet.
14
Stevie
person-098@northwindhealth.example
RevOps Power User
738 msgs · 97 convs · 46d
commits 0 · +0/−0
3.5 Stevie is Northwind Health's most analytically sophisticated AI user, consistently running deep financial-intelligence sessions — G&A benchmarking, AR forensics, payment interchange optimization — that produce board-ready outputs with minimal supervision. His high tool-use volume and rapidly maturing prompt craft make him a strong candidate for a Finance/RevOps AI Lead role, though he has yet to publish reusable assets or visibly pull colleagues into his workflows.
15
Morgan
person-025@northwindhealth.example
Other Power User
528 msgs · 41 convs · 25d
commits 0 · +0/−0
3.5 Morgan is Northwind Health's most prolific HR/Payroll AI user, turning complex 360°/9-Box data and multi-jurisdiction payroll records into polished dashboards and actionable tax-risk analyses with impressive domain depth. She executes autonomously at high volume, but has yet to document, share, or scale her work in ways that would elevate peers around her.
16
Sloane
person-043@northwindhealth.example
Marketing Workflow Embedder
404 msgs · 59 convs · 34d
commits 0 · +0/−0
3.5 Sloane is a marketing operator who has moved well beyond one-off prompts — she is systematizing AI into repeatable production workflows, most notably a self-service HTML tool that generates full post-webinar email sequences each week without manual rework. Her domain fluency (buyer journey framing, compliance verticals, campaign sequencing) gives her outputs real business utility, and her trajectory from copy tasks to API-wired tooling signals strong growth potential as a functional AI lead.
17
Aspen
person-081@northwindhealth.example
Learning Builder
191 msgs · 10 convs · 7d
commits 0 · +0/−0
3.5 Aspen is Northwind Health's sharpest e-learning technologist, demonstrating expert-level command of Rise 360, Storyline, LMS APIs, and Python automation across genuinely complex, production-impacting problems. He drives work end-to-end with strong planning instincts, though zero commits in the window mean his output is not yet verifiably landed in shared codebases.
18
Nolan
person-069@northwindhealth.example
Marketing Workflow Embedder
2,307 msgs · 173 convs · 58d
commits 0 · +0/−0
3.3 Nolan is a high-volume, self-directed channel marketing operator who has woven AI deeply into his daily workflows — from partner sequencing and drip campaigns to automation scripts and formal program documentation. He shows genuine domain depth and creative initiative in scaling his channel function, though the absence of committed, maintainable artifacts means his impact remains harder to verify at the organizational level.
19
Wyatt
person-054@northwindhealth.example
Sales Power User
1,556 msgs · 88 convs · 21d
commits 0 · +0/−0
3.3 Wyatt is a high-volume, self-directed sales rep who has deeply embedded AI into his daily prospecting and call-prep workflows — running automated account searches, building intel decks, and drafting challenger-sales collateral with impressive domain specificity. The main gap is durability: his work lives in chat sessions rather than shared projects or documented processes, limiting its leverage for the broader team.
20
Skyler
person-042@northwindhealth.example
Sales Workflow Embedder
1,336 msgs · 172 convs · 63d
commits 0 · +0/−0
3.3 Skyler is a high-volume Sales practitioner who has deeply embedded AI into his daily deal cycle — turning Gong transcripts into polished executive one-pagers, ROI models, and prospect-specific proposals with remarkable consistency across 63 active days. He's a strong individual executor and fast learner, but has not yet codified his workflows into shared projects or templates that would multiply his impact across the team.
21
Jamie
person-016@northwindhealth.example
Other Workflow Embedder
1,188 msgs · 104 convs · 38d
commits 0 · +0/−0
3.3 Jamie is Northwind Health's most active HR-ops power user, consistently driving AI from prompt to polished deliverable across onboarding automation, vendor evaluation, and employee calibration workflows. With 2,000 tool calls in under two months and rising prompt craft, he is a strong Functional AI Lead candidate for People Ops — his main gap is visible peer influence and committing reusable assets others can build on.
22
person-044@northwindhealth.example
Sales Power User
1,098 msgs · 163 convs · 36d
commits 0 · +0/−0
3.3 is a high-volume Sales rep who has made AI a genuine daily weapon — using it to prep every demo, revive dormant deals, and coach himself off call recordings with real sophistication. His prompt craft is noticeably above average and his output quality is consistent, but impact stays personal; the next step is packaging his playbooks so the broader sales team benefits.
23
Nico
person-026@northwindhealth.example
Other Builder
441 msgs · 78 convs · 45d
commits 0 · +0/−0
3.3 Nico is a hands-on builder who has driven a complex ATS migration (SmartRecruiters → Greenhouse) almost entirely through AI-assisted Python scripting, API automation, and workflow tooling — demonstrating real functional depth in HR/recruiting operations and IT. His breadth of tooling (n8n, Zapier agents, local LLMs, network config) and consistent end-to-end execution make him a strong Functional AI Lead candidate, though the absence of committed code and peer-adoption signals are gaps worth probing.
24
Rowan
person-036@northwindhealth.example
Sales Power User
365 msgs · 25 convs · 11d
commits 0 · +0/−0
3.3 Rowan is a high-volume BDR who has embedded AI deeply into his daily prospecting workflow — mining Salesforce, building talk tracks, and iterating cold outreach to a degree well above his peers. He shows genuine builder instincts (custom SF tools, urgency-sorted widgets, deployable artifacts) but hasn't yet scaled his work into shared assets or demonstrated peer influence.
25
Zion
person-055@northwindhealth.example
Sales Builder
320 msgs · 19 convs · 14d
commits 0 · +0/−0
3.3 Adair is a self-directed sales AE who moved from exploratory SF connector testing to building and deploying a live multi-system prospecting automation within a single month — one of the most concrete tool-building arcs in the cohort. His depth in sales workflows is clear and his speed of adoption is impressive; the next step is externalizing that work into shared projects or commits so teammates can benefit from what he's built.
26
Sasha
person-039@northwindhealth.example
Engineering Workflow Embedder
285 msgs · 36 convs · 24d
commits 0 · +0/−0
3.3 Sasha is Northwind Health's hands-on IT/ops operator who has embedded AI directly into real compliance and access-management workflows — from SOC audit prep to automated onboarding — with impressive self-direction over a short window. His biggest gap is leaving no documented artifacts or commits, which limits his organizational multiplier and keeps his impact personal rather than team-wide.
27
Riley
person-033@northwindhealth.example
Marketing Workflow Embedder
239 msgs · 20 convs · 16d
commits 0 · +0/−0
3.3 Riley is Northwind Health's most productive AI-assisted content producer in Marketing, having systematically used Claude to build a multi-segment sales playbook library, competitive intelligence repository, and interactive demo project plan with strong prompt craft and format fidelity. He thinks ambitiously about AI leverage (digital twin concept) but has not yet crossed into shared tooling or peer-enabling work, making him a high-value individual contributor with untapped multiplier potential.
28
Mika
person-092@northwindhealth.example
Sales Workflow Embedder
603 msgs · 58 convs · 29d
commits 0 · +0/−0
3.2 Mika is a high-activity Sales rep who has embedded AI deeply into her daily prospecting and account management workflows — cross-referencing CMS data, building interactive dashboards, and producing deduplicated outreach lists at scale. She drives strong personal productivity gains but has not yet translated that into shared artifacts or peer enablement that would mark her as a team-level multiplier.
29
Finn
person-086@northwindhealth.example
Sales Workflow Embedder
551 msgs · 115 convs · 27d
commits 0 · +0/−0
3.2 Finn is a high-output Sales rep who has moved quickly from basic prospecting queries to building custom Gong-Salesforce skills, embedding AI directly into his daily pipeline workflow. His functional depth and self-driven experimentation are real strengths, but he leaves no documented or shared artifacts behind — making his impact largely personal rather than team-multiplying so far.
30
Casey
person-005@northwindhealth.example
Marketing Workflow Embedder
429 msgs · 45 convs · 16d
commits 0 · +0/−0
3.2 Casey is Northwind Health's Product Marketing Manager who has quickly made AI a core part of her GTM execution loop — from rebuilding rollout decks to automating newsletter drafts and cross-channel release audits. She's a strong Workflow Embedder with real functional depth, but hasn't yet produced shared artifacts or commits that would signal broader team leverage.
31
Peyton
person-030@northwindhealth.example
Sales Builder
342 msgs · 5 convs · 2d
commits 0 · +0/−0
3.2 Peyton is a self-directed Sales rep who in just two days independently built production-grade AI tooling — a 100+ prospect email pipeline, a cold-call script app, and demo prep utilities — entirely on his own initiative. His output quality and tool fluency are well above the sales peer baseline, though his impact stays personal until he documents and shares his builds with the team.
32
Ryan
person-037@northwindhealth.example
Customer Success Workflow Embedder
237 msgs · 31 convs · 18d
commits 0 · +0/−0
3.2 Ryan is a CS leader who has embedded AI deeply into her day-to-day workflows — surfacing a miscalculated NPS, building churn policy from raw MRR data, and producing board-ready slides quantifying $86K+ in AI-driven savings. Her domain knowledge is strong and her output quality is real, but impact is currently self-contained; the next step is externalizing her work into shared assets and documented processes others can build on.
33
Sutton
person-046@northwindhealth.example
Learning Workflow Embedder
62 msgs · 7 convs · 5d
commits 0 · +0/−0
3.2 Sutton is a high-initiative individual contributor who is rapidly embedding AI into real operational workflows — Jira, SCORM, Gmail invoicing, and Salesforce — demonstrating genuine cross-functional depth. His personal execution is strong, but with zero shared commits or authored projects, his impact remains siloed and has not yet scaled to peers.
34
Phoenix
person-071@northwindhealth.example
Sales Power User
2,086 msgs · 110 convs · 22d
commits 0 · +0/−0
3.0 Phoenix is a high-volume Sales AE who has made AI a near-daily co-pilot for the full deal cycle — discovery prep, stakeholder mapping, demo scripting, CRM logging, and follow-up sequencing across 10+ named accounts in 30 days. His depth of use is impressive, but without documented templates, shared projects, or any artifacts that outlive the chat thread, his impact remains personal rather than organizational.
35
Elliot
person-009@northwindhealth.example
Other Workflow Embedder
2,059 msgs · 79 convs · 28d
commits 0 · +0/−0
3.0 Elliot is Northwind Health's most active legal/paralegal AI user, deeply embedding Claude into contract redlines, dispute communications, and compliance policy work across 79 conversations in under six weeks. He executes complex, multi-turn legal workflows with clear domain fluency, but has yet to formalize that output into shared projects or templates that would extend his impact to peers.
36
Wesley
person-079@northwindhealth.example
Other Workflow Embedder
709 msgs · 54 convs · 18d
commits 0 · +0/−0
3.0 Nataly is HR/Legal's most consistent AI practitioner, embedding Claude into daily document workflows — corrective actions, reference checks, warning notices, and presentations — with notable depth and volume over just 18 active days. She has strong functional grounding and executes reliably, but growth toward true AI leadership will require moving from solo production work to building shared templates and mentoring peers.
37
Pax
person-095@northwindhealth.example
Learning Workflow Embedder
669 msgs · 42 convs · 25d
commits 0 · +0/−0
3.0 Pax is a high-volume, self-directed AI practitioner who has embedded Claude deeply into healthcare competency and skill-assessment content production, generating standardized documents across dozens of clinical roles with minimal oversight. Her domain expertise in SNF, AL, and BH workforce development is clearly strong, but her impact would compound significantly if she shared her templates and workflows with peers and moved toward version-controlled, reusable deliverables.
38
Avery
person-002@northwindhealth.example
RevOps Power User
599 msgs · 55 convs · 24d
commits 0 · +0/−0
3.0 Avery is a high-frequency, self-directed RevOps operator who uses AI fluently across financial modeling, data reconciliation, and strategic tooling research, with a standout 162-message AI P&L session showing real analytical ambition. His ceiling is limited by a lack of documented, shareable artifacts and no visible peer influence — if he starts building things others can use, he becomes a genuine team multiplier.
39
Tatum
person-049@northwindhealth.example
Other Workflow Embedder
498 msgs · 40 convs · 19d
commits 0 · +0/−0
3.0 Tatum is a high-volume, operationally focused AI user who has made AI a genuine daily tool for real HR, payroll, and compliance work — producing polished artifacts like tax shortfall analyses and branded employee guides with notable independence. Her depth is functional and practical rather than exploratory, and her impact appears self-contained; the next step is connecting her outputs to team workflows so her productivity becomes a multiplier for others.
40
Sloan
person-074@northwindhealth.example
Marketing Workflow Embedder
376 msgs · 28 convs · 23d
commits 0 · +0/−0
3.0 Sloan is a high-frequency marketing practitioner who has meaningfully woven AI into her daily copywriting, campaign ideation, and healthcare market research workflows. She demonstrates real domain depth in healthcare compliance and market segmentation, but would amplify her impact significantly by documenting and sharing her prompting approaches with the broader team.
41
Lior
person-091@northwindhealth.example
Sales Power User
208 msgs · 19 convs · 15d
commits 0 · +0/−0
3.0 Lior is a sales rep with genuine credentialing domain knowledge who has made AI a consistent part of his prospecting workflow — building ROI docs, RFP responses, and targeted cold outreach at a pace and depth well above average. His ceiling is limited right now by a lack of documented, shareable outputs and no visible cross-team impact, but the raw usage signal and functional instinct are real.
42
Briar
person-082@northwindhealth.example
RevOps Workflow Embedder
206 msgs · 10 convs · 6d
commits 0 · +0/−0
3.0 Briar is a RevOps practitioner who has quickly embedded AI into real data work — Salesforce reconciliation, list scrubbing, and cross-team content support — with meaningful depth in a short window. She operates hands-on and independently, but has yet to document, share, or scale her workflows beyond her own sessions.
43
Niven
person-093@northwindhealth.example
Marketing Power User
932 msgs · 159 convs · 70d
commits 0 · +0/−0
2.8 Niven is Northwind Health's CMO and a high-volume, self-directed Claude user who leans on AI for executive deliverables — board decks, RFP responses, GTM strategy — with genuine domain fluency. He operates as a capable personal power user but has not yet systematized or socialized his AI practice in ways that multiply impact across his team.
44
Yael
person-080@northwindhealth.example
Sales Workflow Embedder
650 msgs · 36 convs · 20d
commits 0 · +0/−0
2.8 Yael is a Sales/BDR enablement operator who has meaningfully embedded AI into her day-to-day production of scripts, decks, and outreach assets, demonstrating real functional depth in BDR workflows. She is a consistent individual contributor who gets outputs shipped, but has not yet shown the organizational rigor, cross-functional influence, or technical depth needed to lead AI adoption for others.
45
Dakota
person-007@northwindhealth.example
Sales Power User
301 msgs · 14 convs · 12d
commits 0 · +0/−0
2.8 Dakota is a high-volume sales rep who has made AI a daily part of his prospecting motion, generating tailored call scripts, emails, and cadences with strong domain specificity around credentialing workflows. His output quality is real and consistent, but he hasn't yet built reusable assets, shared his methods with peers, or pushed into more structured or creative AI applications.
46
Kendall
person-023@northwindhealth.example
Sales Power User
270 msgs · 29 convs · 18d
commits 0 · +0/−0
2.8 Kendall is a high-volume, output-oriented sales/solutions consultant who reaches for Claude instinctively to produce polished customer-facing deliverables — decks, SLAs, webinar content, and pitch language — across a wide range of deal situations. He gets real work done but hasn't yet structured or shared his AI workflows in ways that compound value for the broader team.
47
Vesper
person-052@northwindhealth.example
RevOps Power User
247 msgs · 21 convs · 13d
commits 0 · +0/−0
2.8 Vesper is a confident RevOps/Sales practitioner who has quickly turned Claude into a personal data analyst — pulling Salesforce pipeline intel, reconciling invoices, and building prospect MRR models with minimal hand-holding. Her impact is real but currently self-contained; the next step is systematizing her workflows into shared projects that multiply her productivity across the team.
48
Justice
person-089@northwindhealth.example
Sales Workflow Embedder
121 msgs · 10 convs · 6d
commits 0 · +0/−0
2.8 Justice is a sales exec actively weaving AI into his day-to-day deal workflow — pulling live Salesforce pipeline data, stress-testing pricing strategy, and building a personal AI 'operating system' to carry persistent sales context. His depth of functional sales knowledge is clear, but he hasn't yet translated that energy into shared artifacts, projects, or commits that would multiply his impact across the team.
49
Drew
person-008@northwindhealth.example
Customer Success Power User
418 msgs · 2 convs · 2d
commits 0 · +0/−0
2.7 Drew is a high-intensity AI user who owns client onboarding and platform training, and has quickly leaned into Claude for both org-design and client-facing copy work across just two sessions. Her depth of functional knowledge is evident, but she has yet to leave a documented, shareable artifact trail — establishing structured outputs and project-level commits would meaningfully amplify her organizational impact.
50
Sage
person-038@northwindhealth.example
RevOps Workflow Embedder
300 msgs · 21 convs · 15d
commits 0 · +0/−0
2.7 Ernesto is a hands-on RevOps practitioner with genuine Salesforce depth — CPQ, SOQL, Apex logs, and data pipeline work — who uses AI as a capable daily co-pilot for technical problem-solving. His impact is currently self-contained: no shared artifacts, projects, or commits suggest his work is flowing back to the team, which is the clearest gap to close.
51
Wells
person-078@northwindhealth.example
Learning Builder
225 msgs · 11 convs · 6d
commits 0 · +0/−0
2.7 Wells is a hands-on builder who quickly moved from basic code translation to constructing a functioning LMS web scraper with merged JSON configs and Excel export — real technical initiative in a short window. She needs to close the loop: committing artifacts, documenting her work, and bringing colleagues along will convert her personal velocity into team-level impact.
52
Onyx
person-094@northwindhealth.example
RevOps Workflow Embedder
169 msgs · 12 convs · 8d
commits 0 · +0/−0
2.7 Onyx is a capable RevOps practitioner who brings solid Salesforce domain depth — DLRS rollups, formula logic, flow automation — and uses AI to accelerate technical problem-solving at a meaningful level. However, with no commits, no authored projects, and limited evidence of documented or shared deliverables, his impact appears contained to individual task execution rather than team-level leverage.
53
Finley
person-011@northwindhealth.example
Sales Power User
85 msgs · 8 convs · 6d
commits 0 · +0/−0
2.7 Finley is a Northeast territory sales rep using AI daily to accelerate prospect research, event targeting, and deal-support collateral with solid domain fluency. He generates real outputs but has not yet moved into structured project ownership or cross-team influence, making him a strong individual contributor with clear upside if pointed toward repeatable, shareable workflows.
54
Kelby
person-090@northwindhealth.example
RevOps Workflow Embedder
81 msgs · 9 convs · 8d
commits 0 · +0/−0
2.7 Kelby is a RevOps practitioner who leans on AI to automate Salesforce data audits and CRM hygiene tasks, demonstrating solid domain fluency in persona classifications, SOQL, and territory mapping. He shows real functional depth but repeatedly restarts similar queries across sessions without consolidating his work, and has yet to deliver a committed, maintainable output — execution discipline is the clearest growth opportunity.
55
Sawyer
person-040@northwindhealth.example
Sales Workflow Embedder
55 msgs · 7 convs · 6d
commits 0 · +0/−0
2.7 Sawyer is a sales rep who has found genuine, repeatable value in AI — using it to produce client-ready ROI documents and Mutual Action Plans tied directly to real deals and Salesforce stages. His functional instincts are strong, but growth is limited by low prompt craft progression, no shared artifacts, and no commit activity; coaching on structured output and template sharing would multiply his impact.
56
Wren
person-053@northwindhealth.example
RevOps Workflow Embedder
624 msgs · 11 convs · 9d
commits 0 · +0/−0
2.5 Wren is a hands-on RevOps operator who has leaned into AI for real data-integration and event-operations work, logging 624 messages and 661 tool calls in under five weeks. Her impact is real but still personal-scale — to become a Functional AI Lead she needs to document her workflows, ship reusable artifacts, and bring peers along.
57
Taylor
person-048@northwindhealth.example
Sales Power User
481 msgs · 78 convs · 36d
commits 0 · +0/−0
2.5 Taylor is a high-volume, self-directed AI user in Sales who reaches for Claude as a fast execution layer — churning out ROI sheets, pricing docs, compliance email sequences, and board briefs with real business context baked in. His output quality is solid and his functional instincts are evident, but incomplete projects, zero commits, and no shared artifacts suggest he's optimizing for personal productivity rather than building durable, team-level leverage.
58
Quincy
person-096@northwindhealth.example
Marketing Workflow Embedder
270 msgs · 28 convs · 19d
commits 0 · +0/−0
2.5 Quincy is a hands-on practitioner who has found real leverage in AI for content QA and operational tasks — her PDF comparison workflows caught dozens of accreditation errors across hundreds of course records. She is currently operating reactively rather than building scalable, documented processes, and would benefit from a push toward project ownership and cross-team knowledge sharing.
59
Joss
person-065@northwindhealth.example
Executive Frontier Explorer
199 msgs · 20 convs · 16d
commits 0 · +0/−0
2.5 Joss is an executive-level explorer who uses AI broadly for strategic research, market intelligence, and personal curiosity — logging 199 messages across 20 sessions in five weeks with strong domain fluency in healthcare UM/CDI and a nose for emerging tech like Agentforce. His impact is currently limited to individual consumption with no shipped artifacts or peer-facing work, making him a high-potential AI champion who still needs to convert exploration into documented, repeatable outputs others can build on.
60
Devon
person-084@northwindhealth.example
Engineering Power User
191 msgs · 14 convs · 11d
commits 0 · +0/−0
2.5 Devon is an IT-department power user who consistently reaches deep, tool-heavy sessions — most notably a high-volume Jira analytics build — demonstrating real functional grounding and willingness to experiment. To move from capable individual contributor to AI Lead, he needs to shift from one-off outputs to documented, repeatable artifacts and begin pulling colleagues along with him.
61
Adair
person-056@northwindhealth.example
Marketing Power User
110 msgs · 23 convs · 18d
commits 0 · +0/−0
2.5 Adair is an active, self-directed AI user who applies Claude across marketing analytics, RevOps data pulls, and lightweight frontend/backend tasks with genuine curiosity. He has not yet translated that individual usage into shared artifacts, documented workflows, or peer influence, so his organizational impact remains largely invisible.
62
Tristan
person-051@northwindhealth.example
Sales Workflow Embedder
98 msgs · 15 convs · 13d
commits 0 · +0/−0
2.5 Jeffery is a Client Executive who has made AI a reliable daily tool for drafting prospect emails, building pricing proposals, and polishing client communications — showing solid domain knowledge of Northwind Health's credentialing and LMS pricing. He is consistent and engaged, but usage remains narrowly transactional with no evidence of building reusable assets, sharing work with peers, or driving multi-step projects to completion independently.
63
Alex
person-001@northwindhealth.example
RevOps Frontier Explorer
74 msgs · 3 convs · 3d
commits 0 · +0/−0
2.5 Alex is a RevOps operator with clear domain knowledge of CRM triggers and subscription workflows who is actively pushing into AI-assisted automation — her 60-message, 21-tool-block email/Slack summarization session signals real ambition. She needs to close execution loops and produce shareable artifacts to move from explorer to builder.
64
Echo
person-085@northwindhealth.example
Sales Workflow Embedder
70 msgs · 11 convs · 9d
commits 0 · +0/−0
2.5 Echo is a solution consultant actively embedding AI into his daily sales prep — demo talk tracks, click paths, and competitive cheat sheets — with a clear trajectory of improving prompt sophistication over 30 days. He has not yet converted this personal productivity gain into documented, shareable assets or team-visible output, which limits his organizational impact today.
65
Sydney
person-047@northwindhealth.example
Sales Workflow Embedder
30 msgs · 3 convs · 3d
commits 0 · +0/−0
2.5 Sydney is a sales rep using AI tactically to accelerate prospect research and call script generation, with visible improvement in prompt quality across a short window. He has real potential as a functional AI lead but needs to move beyond one-off tasks toward building reusable assets and demonstrable execution to earn that role.
66
person-068@northwindhealth.example
Sales Workflow Embedder
126 msgs · 10 convs · 4d
commits 0 · +0/−0
2.3 Nolan is a sales rep who has meaningfully embedded AI into his day-to-day prospecting and call-summary workflows, with a notable 92-message deep dive into competitive positioning against Northwind Health. His usage is consistent and functional but remains narrow and transactional — he has not yet moved into building reusable assets, coaching peers, or driving broader AI adoption.
67
Joss
person-064@northwindhealth.example
Sales Power User
110 msgs · 9 convs · 7d
commits 0 · +0/−0
2.3 Joss is an active Sales team member who consistently turns to AI for practical CRM and pipeline tasks — Salesforce queries, conference list enrichment, and call-to-email workflows — showing real functional intent. His usage is genuine but remains transactional and task-by-task; to become a Functional AI Lead he needs to move from ad-hoc queries to documented, repeatable workflows others can adopt.
68
Linden
person-066@northwindhealth.example
Sales Workflow Embedder
94 msgs · 17 convs · 3d
commits 0 · +0/−0
2.3 Linden is a sales rep who has quickly embedded AI into daily prospecting, generating CRM-informed call scripts at volume across a range of healthcare personas including CHROs, compliance directors, and HR managers. The usage pattern is consistent and productivity-oriented but remains shallow — no projects, no commits, and no evidence yet of building reusable assets or influencing peers.
69
Indigo
person-015@northwindhealth.example
Sales Workflow Embedder
50 msgs · 5 convs · 4d
commits 0 · +0/−0
2.3 Indigo is an early-stage AI adopter using Claude primarily to accelerate day-to-day sales tasks — prospect emails, talk tracks, and upsell prep — with solid functional grounding in Northwind Health's customer workflows. She's getting real value from the tool but has not yet moved beyond delegating individual tasks to building reusable assets or influencing peers.
70
Oakley
person-029@northwindhealth.example
Other Workflow Embedder
505 msgs · 12 convs · 11d
commits 0 · +0/−0
2.2 Oakley is an active but early-stage AI user who has leaned on the tool primarily for operational and administrative tasks — branded menus, expense tracking, vendor disputes — rather than Northwind Health core workflows. With 505 messages across 11 days she shows genuine engagement, but without project authorship, commits, or domain-specific depth, she is not yet demonstrating the functional AI leadership profile the program is targeting.
71
Reese
person-032@northwindhealth.example
Sales Power User
373 msgs · 36 convs · 16d
commits 0 · +0/−0
2.2 Reese is an active BDR who uses AI consistently as a daily email-drafting assistant, logging 373 messages across 30 days of prospecting work. His usage is high-volume but narrow — he has not yet moved beyond transactional copy generation into building reusable assets, structured workflows, or peer-facing outputs.
72
Kendall
person-019@northwindhealth.example
Executive Frontier Explorer
106 msgs · 10 convs · 6d
commits 0 · +0/−0
2.2 Kendall shows curiosity and creative energy — experimenting with AI for pitch decks, visual design, and financial concepts — but usage skews toward personal and side-project work rather than Northwind Health workflows. Without commits, documented outputs, or evidence of departmental impact, he has not yet converted AI interest into measurable organizational value.
73
Hayden
person-013@northwindhealth.example
Other Frontier Explorer
1,355 msgs · 84 convs · 44d
commits 0 · +0/−0
2.0 Hayden is a high-volume, curious AI user who has developed real fluency with complex, multi-turn sessions — but virtually all of that activity is directed at personal academic work (architecture and engineering coursework) rather than Northwind Health workflows. As an AI Lead candidate, there is currently no evidence of functional domain expertise, work deliverables, or peer influence that would translate to organizational impact.
74
person-010@northwindhealth.example
Marketing Workflow Embedder
74 msgs · 2 convs · 2d
commits 0 · +0/−0
2.0 is a marketing intern on the demand-gen team who used AI to power through a structured prospect-research task, demonstrating early but promising engagement with AI-assisted workflows. With only 2 active days, 2 conversations, and no commits or project artifacts, there is insufficient data to assess her ceiling — she warrants a 30-day revisit once she has more reps.
75
Cleo
person-058@northwindhealth.example
Other Power User
32 msgs · 12 convs · 7d
commits 0 · +0/−0
2.0 Cleo is an occasional AI user who reaches for Claude when he needs quick legal or contractual answers, spanning both personal matters and work topics like license terms and HR compliance. With no commits, no projects, and shallow conversation arcs, he has not yet demonstrated the depth, structure, or execution habits needed to function as a Functional AI Lead.
76
Crew
person-083@northwindhealth.example
Sales Power User
732 msgs · 45 convs · 14d
commits 0 · +0/−0
1.8 Crew is an active daily user who has found AI useful for transactional sales tasks like pricing doc creation and client email polish, but his usage skews personal and lacks depth, structure, or delivery artifacts. To become a functional AI lead he would need to move from ad-hoc prompting toward documented, repeatable workflows with measurable business outcomes.
77
Jordan
person-017@northwindhealth.example
Sales Power User
54 msgs · 5 convs · 3d
commits 0 · +0/−0
1.8 Jordan is an early-stage AI adopter in Sales using Claude primarily for reactive prospect research and ICP qualification checks. With only 3 active days, no structured outputs, and no commits, he has not yet demonstrated the depth, organization, or drive needed for a Functional AI Lead role.
78
Kai
person-018@northwindhealth.example
Sales Power User
26 msgs · 5 convs · 5d
commits 0 · +0/−0
1.8 Kendall is an early-stage AI user applying Claude primarily to sales prospecting tasks such as account research and engagement prioritization, with limited depth or follow-through visible so far. With only 5 conversations, no authored projects, and no commits over a two-week window, there is insufficient evidence to assess most leadership attributes; he warrants a check-in to understand whether higher-value work is happening outside this tool.
79
Rune
person-097@northwindhealth.example
Sales Power User
18 msgs · 2 convs · 2d
commits 0 · +0/−0
1.8 Rune is a Sales rep making early, transactional use of AI for call analysis and prospect follow-up emails — a practical start but not yet showing depth, initiative, or creative extension of the tool. With only 2 conversations and no built artifacts, there is insufficient data to assess his ceiling; worth revisiting in 30 days if usage grows.
80
Greer
person-087@northwindhealth.example
Learning Workflow Embedder
12 msgs · 1 convs · 1d
commits 0 · +0/−0
1.8 Greer shows early promise applying AI to regulatory content QA, successfully identifying catalog discrepancies across PDF versions in his first session. With only one active day and no commits, there is far too little data to assess sustained impact or growth trajectory.
81
Sterling
person-045@northwindhealth.example
Learning Workflow Embedder
57 msgs · 5 convs · 4d
commits 0 · +0/−0
1.7 Sterling shows early, task-oriented AI usage — primarily compliance lookups and content copy edits — but has not yet demonstrated structured, self-directed, or cross-functional AI work. With only 4 active days and no authored projects or commits, she is at the awareness stage and would benefit from structured onboarding to unlock deeper impact.
82
Oslo
person-070@northwindhealth.example
Marketing Workflow Embedder
14 msgs · 5 convs · 4d
commits 0 · +0/−0
1.7 Oslo shows early-stage AI exploration, primarily using Claude for visual/marketing asset generation and a single compliance document review, but with only 14 messages across 5 sessions and no project artifacts or commits, meaningful impact is not yet evident. She has a foundation to grow from — particularly in marketing applications — but needs to move from isolated experiments to repeatable, documented workflows to qualify as a Functional AI Lead candidate.
83
Phoenix
person-072@northwindhealth.example
Customer Success Power User
6 msgs · 1 convs · 1d
commits 0 · +0/−0
1.7 Phoenix has one recorded AI session focused on a Salesforce account analysis that surfaced a churned $7,500 MRR customer — a useful, real-world query, but far too limited to assess sustained capability. With only a single active day and no projects or commits, there is insufficient signal to evaluate him as a Functional AI Lead candidate at this time.
84
Gray
person-062@northwindhealth.example
Customer Success Workflow Embedder
10 msgs · 3 convs · 2d
commits 0 · +0/−0
1.5 Gray is in the earliest stage of AI adoption, with only 10 messages across 2 days and no completed workflows or commits to evaluate. Current usage is exploratory and task-oriented (email drafting, Salesforce lookup), with insufficient data to assess most functional AI lead attributes meaningfully.
85
Ellis
person-060@northwindhealth.example
RevOps Methodical Operator
6 msgs · 1 convs · 1d
commits 0 · +0/−0
1.3 Ellis has made only one appearance — a single NetSuite how-to query — providing almost no signal on AI capability or initiative. Insufficient data to assess meaningfully; re-evaluate if engagement increases.
86
Tate
person-076@northwindhealth.example
Other Power User
18 msgs · 4 convs · 2d
commits 0 · +0/−0
1.2 Tate has only 18 messages across 2 active days, with usage limited to a personal best-man speech and a vague employee-count lookup — no work-domain AI application is visible. There is insufficient signal to assess functional expertise, organized output, or peer influence at this time.
87
Ira
person-088@northwindhealth.example
Sales Power User
8 msgs · 3 convs · 1d
commits 0 · +0/−0
1.2 Ira has made only the earliest exploratory contact with the AI platform — three brief sessions on a single day, none of which produced a working output or demonstrated domain depth. At this stage there is insufficient signal to assess potential; meaningful re-evaluation requires sustained engagement over multiple weeks.
88
Adair
person-057@northwindhealth.example
Other Methodical Operator
4 msgs · 1 convs · 1d
commits 0 · +0/−0
1.2 Adair has only one recorded session with a single exploratory prompt that never gained traction, providing almost no basis for meaningful evaluation. There is insufficient evidence of expertise, follow-through, or AI fluency at this stage.
89
Remy
person-073@northwindhealth.example
Learning Workflow Embedder
4 msgs · 2 convs · 2d
commits 0 · +0/−0
1.2 Remy is a very early-stage user with only 4 messages across 2 sessions, focused on basic onboarding and tool access rather than applied AI work. Insufficient activity exists to assess meaningful capability; she should be revisited after at least 30 days of active use.
90
Tate
person-075@northwindhealth.example
Other Power User
2 msgs · 1 convs · 1d
commits 0 · +0/−0
1.2 Tate has only one recorded session with a single personal-interest prompt unrelated to Northwind Health workflows, providing virtually no signal on AI capability or work application. Insufficient data to assess meaningful impact; engagement and work-relevant use cases need to develop before any evaluation is possible.
91
Cameron
person-004@northwindhealth.example
Sales Power User
2 msgs · 1 convs · 1d
commits 0 · +0/−0
1.0 Cameron has one recorded interaction — a basic lookup query about free Northwind Health resources for hospitals — which provides virtually no basis for evaluation. Insufficient data exists to assess any of the six AI Lead attributes at this time.
06 · Functional AI Lead candidates

Marketing's lead is already shipping. Sales and CS still need theirs.

Lane has been Marketing's federated AI lead since March. He's shipped a working multi-agent pipeline, a lead-verification gate that's already running BDR, and a content-tagging engine. The card below shows what he built and what to do with it. Sales and CS don't have a counterpart yet — the candidate pools beneath surface the people closest to ready.

Open roles — candidate pools
Sales
#1
Logan
person-022@northwindhealth.example
Logan is the only sales rep whose summary explicitly notes he has already pushed Claude access to peers, and he uniquely scores a 3 on trusted_peer where the rest of the pool sits at 2. Across 50 active days and 124 conversations he's consistently turning prospect intel and pricing data into exec-ready deliverables (e.g., 'consolidate this into three pages... add some parts about the compliance risk'), making him the closest match to the 'trusted peer whose adoption inspires others' criterion.
Risk: Only 1 project built so far — needs to demonstrate he can codify his workflows into maintainable, shared assets.
Recommended next step: Within 14 days, have Logan package his top 3 prospect-intel and pricing prompts into a shared Sales project and demo it at a sales team standup.
#2
Robin
person-035@northwindhealth.example
Riley is the strongest raw talent in the pool — the only sales candidate scoring 5/5 on creative_problem and learns_fast, with around 1,000+ messages and a couple thousand tool calls in 50 days producing pricing guides, competitive business cases, and negotiation decks. His quote 'I want to teach you everything about my work style... share with you my numbers so that you know exactly what it will take' shows the systems-thinking instinct of a functional lead.
Risk: Zero projects and trusted_peer score of 2 — currently a lone wolf; needs to prove he'll evangelize, not just execute.
Recommended next step: Ask Riley to convert his Author Health custom quote workflow into a documented, reusable competitive-deal template and present it to two peers by end of month.
#3
Frankie
person-012@northwindhealth.example
Frankie already has 4 projects built (the highest project count of any sales candidate) and demonstrates exceptional domain depth — her RFP work and FQHC-vs-Healthstream competitive positioning ('We need to stand out here as the platform that is built specifically for FQHCs') show she thinks at the market-segment level, not just deal level. The combination of deep_expertise 5 and existing project hygiene makes her the most 'organized driver' profile in sales.
Risk: Lower volume (53 convs, 22 active days) than peers — need to confirm she has bandwidth to take on a cross-team lead role.
Recommended next step: Schedule a 30-min scoping call to assess her capacity and have her share her 4 existing projects with Joss to evaluate maintainability.
Customer Success
#1
Ryan
person-037@northwindhealth.example
Ryan is the only viable CS lead in the pool — she's already operating at a leadership altitude, having surfaced a miscalculated NPS, built churn policy from raw MRR data, and produced board-ready slides quantifying $86K+ in AI-driven savings. Her 18 active days across 31 conversations show sustained, embedded use, and her scores (4/4/3/3/2/3) are the most balanced in CS with real driver and creative_problem signal.
Risk: Trusted_peer score of 2 and zero projects — needs to externalize her work and build visible peer influence to earn the 'Lead' title.
Recommended next step: Within 14 days, ask Ryan to convert her $86K savings analysis into a CS-team-wide project template and present her NPS methodology fix at the next CS staff meeting.
07 · Top projects

The work that compounded — projects with knowledge written in

A project with docs attached is a person who took the time to write context their teammates can reuse. That is the single move that makes an AI workflow repeatable. These twelve projects are the closest Northwind Health has to a starter library; Tiger Team should freeze them, document them, and hand a copy to every new Functional AI Lead on day one.

Sales Owl
River · 7 docs · 154 convs by creator
Internship Project
Jamie · 3 docs · 104 convs by creator
Orthopedic Practices
Riley · 3 docs · 20 convs by creator
Ambulatory Surgery Centers Market Segment
Riley · 2 docs · 20 convs by creator
Behavioral Healthcare Market Segment
Riley · 2 docs · 20 convs by creator
P&L Improvement From AI
Avery · 2 docs · 56 convs by creator
Northwind Health Elevator Pitch
Riley · 1 docs · 20 convs by creator
How to use Claude
River · 1 docs · 154 convs by creator
An example project that also doubles as a how-to guide for using Claude. Chat with it to learn more about how to get the most out of chatting with Claude!
Virtual Twin
Nolan · 1 docs · 174 convs by creator
Urgent Care Market Segment
Riley · 1 docs · 20 convs by creator
I am trying to better understand the market dynamics in the Urgent Care segment in order to position Northwind Health for maximum effectiveness.
Post-webinar sequence generator
Sloane · 1 docs · 59 convs by creator
Product Marketing Charter
Riley · 1 docs · 20 convs by creator
Define and document a charter statement for Product Marketing at Northwind Health.
08 · What the team uses AI for

Theme map across all conversations

Sales prospecting is the single largest theme by a wide margin. RevOps data work is second. That ratio is your real Q3 priority signal for the Tiger Team. The named contributors on each row are the in-house experts for that domain — if Steve needs an opinion on AI in CS, Ryan is the person; on Compliance, Hollis; on Marketing, Niven.

Sales Prospecting
971 (23.8%)
person-031 (137) · person-003 (128) · person-071 (66)
Other
495 (12.2%)
person-009 (66) · person-061 (53) · person-002 (43)
Revops Data
343 (8.4%)
person-061 (74) · person-034 (45) · person-031 (42)
Sales Call Analysis
287 (7.0%)
person-003 (51) · person-086 (41) · person-044 (36)
Personal Non Work
266 (6.5%)
person-093 (39) · person-069 (36) · person-013 (26)
Writing Communication
248 (6.1%)
person-061 (74) · person-069 (24) · person-016 (13)
Marketing Campaigns
199 (4.9%)
person-043 (50) · person-069 (23) · person-074 (20)
Hr Talent
188 (4.6%)
person-079 (27) · person-016 (25) · person-042 (23)
Ai Tooling
168 (4.1%)
person-034 (25) · person-022 (17) · person-003 (17)
Compliance Policy
156 (3.8%)
person-077 (67) · person-028 (21) · person-014 (20)
Sales Pricing
125 (3.1%)
person-061 (22) · person-042 (20) · person-035 (12)
Operations
111 (2.7%)
person-039 (17) · person-026 (14) · person-005 (9)
Research Market Intel
97 (2.4%)
person-093 (23) · person-014 (11) · person-033 (6)
Finance Pricing Analysis
89 (2.2%)
person-098 (30) · person-025 (9) · person-049 (9)
Customer Success
86 (2.1%)
person-037 (13) · person-003 (10) · person-092 (7)
Learning Course Authoring
59 (1.4%)
person-095 (10) · person-049 (5) · person-016 (4)
Executive Strategy
46 (1.1%)
person-093 (10) · person-069 (5) · person-061 (5)
Automation Scripts
36 (0.9%)
person-026 (9) · person-016 (6) · person-039 (4)
Engineering Frontend
32 (0.8%)
person-096 (13) · person-056 (6) · person-034 (4)
Prompt Engineering
28 (0.7%)
person-069 (8) · person-034 (5) · person-086 (4)
Data Analysis Sql
24 (0.6%)
person-034 (3) · person-056 (2) · person-096 (2)
Engineering Backend
18 (0.4%)
person-026 (4) · person-034 (3) · person-081 (3)
09 · Recurring use cases

Where people are repeating the same work

Each card below is a task pattern that 4 or more conversations across multiple people fit. These are the highest-leverage Tiger Team productization targets — every conversation that fits one of these patterns is work the team could be doing with a starter skill instead of from scratch. Ranked by total conversations × unique contributors.

Drafting prospect outreach and follow-up emails
203 conversations · 29 people
Reps write, refine, or polish cold outreach, warm outreach, follow-up, and re-engagement emails to healthcare prospects.
"Drafted a low-pressure follow-up email for stalled prospect Gabby at Nira Medical, referencing deal context and board meeting delays."
Drafting and polishing professional emails and messages
98 conversations · 27 people
Users ask Claude to draft, reword, or polish outgoing emails, Slack messages, and customer-facing communications for tone and clarity.
"Drafted client email navigating billing dispute while protecting signed contract terms and avoiding clawback risk."
Querying Salesforce pipeline and account data
86 conversations · 27 people
Users query Salesforce (often via MCP connector) to pull pipeline totals, opportunity status, account/contact lists, or build targeted ABM lists.
"Retrieved Q2 2026 Salesforce pipeline: $43.1M across 851 opps, broken down by sales stage."
Installing and troubleshooting Claude tooling
65 conversations · 34 people
Users install, configure, and troubleshoot Claude Code, Cowork, Desktop, and connectors to third-party tools like Salesforce and Outreach.
"User asked how to install Claude Code; assistant provided platform-specific installation instructions."
Researching prospect accounts before outreach or calls
111 conversations · 19 people
Reps look up org details, contact info, employee counts, classifications, and pain signals for a specific prospect to inform outreach or call prep.
"Researched Northwest Kidney Centers leadership contacts and clinic details for sales prospecting outreach."
Extracting CRM notes and summaries from call transcripts
74 conversations · 20 people
Reps paste a sales call transcript (often from Gong) to generate structured CRM-ready notes, summaries, and Salesforce field updates.
"AE submitted a sales call transcript and received structured CRM-ready notes with key next steps and deal status."
Building pre-call discovery and demo prep briefs
70 conversations · 19 people
Reps synthesize CRM, Gong, and web data into a structured briefing document with discovery questions before an upcoming sales call or demo.
"Generated SC demo prep brief for Wellness MD Group prospect from copied call transcript."
Building pricing proposals and negotiation responses
50 conversations · 20 people
Reps construct pricing proposals, quote comparisons, objection rebuttals, and negotiation strategies for active deals.
"Built a branded Google Doc comparing 3 renewal pricing options with savings vs. auto-renewal for a VP of Revenue."
Drafting LinkedIn posts and marketing copy
32 conversations · 13 people
Writing LinkedIn post copy and marketing content to promote Northwind Health webinars, product features, and blog content.
"Generated multiple LinkedIn post options to promote a Northwind Health credentialing webinar registration."
Cold call script generation for prospects
33 conversations · 9 people
Reps generate personalized cold call scripts tailored to specific healthcare personas using CRM and persona data.
"Generated a tailored cold call script for a CHRO prospect at a specialty oncology healthcare organization."
Analyzing call transcripts for deal coaching
29 conversations · 10 people
Sales coach personas are applied to call transcripts to assess deal viability, buyer signals, and produce coaching recommendations.
"Rep used a richly engineered coach persona to extract deal viability, buyer roles, and a prioritized feature requirements breakdown from a Gong call transcript."
Building prospect-facing executive summary documents
24 conversations · 8 people
Reps build polished one- to two-page executive summary or ROI PDFs tailored to a specific prospect after a demo or discovery call.
"Built a custom ROI PDF for Horizon Health credentialing transition using real quote data from Q-31588-1 for 50 providers."
10 · Best prompts gallery

Twelve prompts the team should be sharing already

Every prompt below was written by a Northwind Health employee in the window, produced a result the LLM judged a breakthrough, and is portable enough to hand to a colleague tomorrow. Put them in an internal channel or a Claude project. The Functional AI Leads should be expected to publish a prompt a week of this quality; it is the lowest-cost adoption mechanism in the federated model.

Oakley compliance_policy
Hola Claude, necesito que me ayudes a revisar y comparar que la traduccion de la parte de procedimiento sea fiel al documento original de inglés. al ser procedimientos, necesito que sean fieles a los descritos en ingles que sean del mismo sentido, que si alguien en español los lee haga y llegue al mismo resultado que el de ingles. Si tienes alguna observacion o correcion necesito que me des una explicacion de porque esta mal, la propuesta del texto corregido e indicarme exactamente en que parte del procedimiento se encunetra Esta es mi propuesta: Expliacion de que esta mal Quote en ingles Quote en español Propuesta de texto corregido Show less 13:59 Instrucciones para Revisión de Traducción de Procedimientos Document  Entendido, aplicaré exactamente ese formato en todas las revisiones. Para confirmar que lo tengo claro: * Cada observación sigue el esquema: Ubicación → Explicación → Quote inglés → Quote español → Texto propuesto * Solo las observaciones del resumen final llevan 🔴 CRÍTICA * Lo que no tiene problema va como ✅ Sin observaciones ¿Continuamos con el siguiente bloque de documentos?
Defines goal (translation fidelity), output schema (4 columns), quality criteria, and severity flags—reusable template for any bilingual procedure review.
Riley marketing_campaigns
i have attached two files for your reference. The first, is a Market Segment Overview synthesized from 5 different market research documents. The second is our Sales Playbook. Inside the Sales Playbook are 8 pages of content for 4 market segments (two pages each); see pages 80-87. Using the Market Segment Overview content and referring to the 2-page example in the Playbook, please construct content for the Urgent Care market segment that matches the style, tone, word-count, and format from the four 2-page examples in the Playbook. Act as a Product Marketing leader at Northwind Health, utilizing any current knowledge about our solutions, much of which can be found in the remaining pages of the Sales Playbook I have attached.
Clear role, explicit source files, format/style/length constraints, and output structure mirror defined examples—strong reusable template.
Blair sales_call_analysis
VIP Community Services prep and post meeting review. You are the most experienced B2B sales coach there has ever been, with a passion for ‘hard truth’ coaching and writing the perfect follow up emails. You don’t beat around the bush with social niceties – you prefer to get straight to the point. When reviewing a sales call transcript, you seem to have a sixth sense intuition for understanding the following things: 1. Whether or not the buyer is serious about their intention or ability to make a purchase in the next few months. 2. Who are the attendees on the client side and their roles. 3. Whether the buyer has expressed a genuine problem that Northwind Health is equipped to solve for them. 4. Whether the prospect’s company has a organizational goal or that ensures budget is being allocated to this 5. Whether the contact(s) came across as a “coach”, “champion”, “opponent”, or “economic buyer”. (label multiple if multiple participants joined from the prospect organization) 6. Finally, whether the rep should progress that call forward and stick with it, or if they should seriously reconsider whether it’s worth their time and energy to keep it in their pipeline. Analyze this call transcript and help coach this rep through their next steps: [insert transcript of Gong call] In your response, include the following sections: 1. If you were the rep, would you keep this deal in your pipeline? Why or why not? What would have to be true for you to stick with it? 2. What does your gut tell you about this contact’s willingness and authority to buy something soon? Is there someone else we need to meet or equip our contact to sell to internally? 3. All action items captured as next steps by either side (Format: Who? What? By when?) 4. Write a concise and professional follow up email that you recommend the rep use to end-cap the call event and mobilize appropriate next steps. Return only the requested information. No acknowledgement of the request. No summary of the actions t
Strong role, sharp 6-point analysis rubric, structured output sections, clear persona—reusable template needing only transcript swap
Tobin compliance_policy
please review the following SMB General policies for full federal compliance, and add or modify them as needed to achieve it, procedures must be expanded to contain operational details and NO codes nor best practices, just explain what what the responsible position does to comply with the requirements, if no responsible position federally mandated is found, use "The facility administrator or designee when individual", they must also be in present tense and have 3 subsections each minimum but not limited to, also all in Arial 11
Clear goal, explicit constraints (no codes, present tense, role fallback, font), structural rules (3 subsections min), reusable compliance template
Nolan sales_call_analysis
Act as a live coaching partner to help me become an industry expert in healthcare compliance, credentialing, and Northwind Health’s core solutions. Context: - I currently lead training calls but rely too heavily on scripts and slides. - My goal is to deeply understand the material so I can confidently go off-script, handle questions, and explain concepts clearly in real-world terms. - I will be using this during my commute, so keep things conversational and interactive. Format: - Ask me one realistic question or scenario at a time (like a partner/customer would ask). - Wait for my response. - Then critique my answer: - What I got right - What I got wrong or unclear - How to improve it to sound like an expert - Then give me a stronger version of the answer. - Keep pushing me with follow-up questions that test deeper understanding (the “why” behind things). Focus areas: - Compliance (HIPAA, OSHA, audits, policy management) - Credentialing (provider verification, expirations, risk) - How Northwind Health’s platform actually solves real operational problems - Clear differentiation between similar concepts (like compliance vs credentialing) Goal: Turn my memorized talking points into real understanding so I can confidently explain, adapt, and go off-script in live trainings.
Clear role, goal, constraints, and iterative format. Reusable template for any domain expertise coaching session—just swap focus areas.
Quinn sales_prospecting
I need you to generate a targeted lookalike prospecting list for Nolan Foster on the Northwind Health BDR team. Please follow these exact steps using our connected Salesforce (MT SF) integration: STEP 1 — Find the BDR in Salesforce Search for the BDR's User record by name: * Object: `User` * Fields: `Id`, `Name`, `Email`, `Title`, `IsActive` * Filter: `Name LIKE '%[BDR NAME]%'` AND `IsActive = true` Capture their Salesforce User ID for use in all subsequent queries. STEP 2 — Pull Their Expo-Completed Accounts from Last Month Query Opportunities where this BDR set and completed the expo last month: * Object: `Opportunity` * Fields: `Id`, `Name`, `Account.Name`, `Account.Id`, `Account.Industry`, `Account.Segment__c`, `Account.BillingState`, `Account.NumberOfEmployees`, `Products_Interested_In_NEW__c`, `LeadSource`, `Opp_Demo_Completed_Date__c` * Filter: `Demo_Set_By__c = '[BDR SALESFORCE ID]'` AND `Opp_Demo_Completed__c = true` AND `Opp_Demo_Completed_Date__c = LAST_MONTH` * Order by: `Opp_Demo_Completed_Date__c DESC` * Limit: 100 From these results, extract: * Unique Industries represented * Unique BillingStates represented * Employee size range (min/max) * Most common Products Interested In STEP 3 — Find Lookalike Accounts (No Demo, No Activity > 14 Days) Using the industries and states identified in Step 2, query for lookalike accounts: * Object: `Account` * Fields: `Id`, `Name`, `Industry`, `Segment__c`, `BillingState`, `NumberOfEmployees`, `LastActivityDate`, `Day_Since_Last_Outreach_Activity__c`, `Outreach_Activity_Greater_Than_14_Days__c`, `Phone`, `Website`, `AccountSource` * Filters: * `Demo_Completed__c = false` * `Outreach_Activity_Greater_Than_14_Days__c = true` * `Industry IN ('[LIST OF INDUSTRIES FROM STEP 2]')` * `BillingState IN ('[LIST OF STATES FROM STEP 2]')` * `NumberOfEmployees >= [MIN FROM STEP 2]` * `NumberOfEmployees <= [MAX FROM STEP 2]` * `Open_Opportunities__c = 0` * Order by: `Day_Since_Last_Outreach_Activity__c DESC` * Limit:
Multi-step Salesforce query template with explicit objects, fields, filters, and dynamic variable slots—highly reusable across BDR team with minimal edits.
sales_call_analysis
You are the most experienced B2B sales coach in the world, known for brutal honesty, pattern recognition, and writing follow-up emails that actually move deals forward. You don't soften hard truths. You don't waste words on pleasantries. When you review a sales call transcript, you cut straight to what matters. Analyze the following call transcript and deliver your coaching response in exactly the sections listed below. No preamble. No acknowledgment of the request. No summary of what you did. **Transcript:** """ [] """ --- **Your analysis must cover the following, in this order:** **1. Pipeline Decision** Would you keep this deal in your pipeline if you were the rep? Give a direct yes or no and the reasoning behind it. If yes, what has to remain true for it to stay. If no, what would have to change before it deserves another minute of the rep's time. **2. Buyer Signals & Authority Assessment** What does your gut tell you about this contact's genuine willingness and actual authority to buy something in the next few months? Call out any red flags or green lights. Is there someone else who needs to be in the room — and if so, who and why? **3. Attendee Profiles** List each attendee from the prospect's organization. For each person, identify their role and label them as one or more of the following: **Coach**, **Champion**, **Opponent**, or **Economic Buyer**. Be specific about what in the transcript led you to that label. **4. Problem & Budget Fit** Has the buyer articulated a genuine problem that Northwind Health is positioned to solve? Is there an organizational goal or strategic initiative that suggests budget is — or will be — allocated to address it? State plainly if either is unclear or missing. **5. Action Items** Every next step committed to by either side, formatted as: - **Who** | **What** | **By When** **6. Follow-Up Email** Write the exact follow-up email the rep should send. It should be concise, professional, and do two things: close the loop on the call and d
Strong role, explicit section structure, clear output format, reusable template—just swap transcript and company name. Minor truncation at end is negligible.
Logan revops_data
You are an expert Salesforce CRM analyst and sales operations consultant with deep knowledge of Salesforce pipeline management, opportunity tracking, and reporting structures. I am a sales leader who wants to compare my team's pipeline against my counterpart Lior's team pipeline, with a specific focus on the **products associated with each team's pipeline opportunities**. Using your Salesforce expertise, help me conduct this comparison by: - Identifying the key Salesforce objects, fields, and reports relevant to comparing pipeline products across two sales teams (e.g., Opportunities, Products/PricebookEntries, OpportunityLineItems, Owner hierarchies) - Outlining what a side-by-side pipeline product comparison between my team and Lior's team would look like — including metrics such as product mix, deal value by product, product frequency, and stage distribution by product - Explaining how this data would typically be surfaced in Salesforce (standard reports, custom report types, dashboards, or SOQL queries) - Highlighting any common gaps or blind spots sales leaders overlook when comparing pipeline product composition across teams - Providing guidance on how to interpret differences in product distribution between the two teams and what those differences might signal about sales strategy, coverage, or performance Assume I have access to Salesforce and standard sales leader permissions. Where relevant, suggest specific Salesforce report types or dashboard configurations I could use to pull this comparison directly.
Clear role, scoped goal, explicit sub-tasks, actionable constraints (permissions, SF objects), reusable for any two-team pipeline comparison
Phoenix prompt_engineering
AI Operating Playbook – Phoenix / Northwind Health (Claude Optimized) 1. Role & Operating Expectation You are my strategic business partner, not just an assistant. Your role is to: * Help me think clearly and make better decisions * Turn rough ideas into structured, high-impact communication * Improve positioning, messaging, and deal strategy * Identify gaps, risks, and missed opportunities Do not default to being agreeable. Your value comes from improving my thinking, not validating it. 2. Communication Style (Strict Rules) * Be concise and structured * Avoid fluff, filler, or generic statements * Prioritize clarity, action, and business impact * Use bullets or sections when possible * If something is unclear or weak → call it out directly When appropriate: * Challenge assumptions * Offer a better alternative * Explain why something should change 3. Default Behavior (Critical for Claude) Claude tends to be polite and passive—override that: * If I give vague input → clarify AND propose a strong version * If I ask for feedback → be direct, not diplomatic * If something is weak → rewrite it better without asking first * If multiple approaches exist → recommend one and explain why Do not say: * “Here are a few options” (unless I ask) Instead: * Give the best option first, then alternatives if needed 4. Business Context * I work at Northwind Health (healthcare compliance platform) * My focus: * Revenue growth * Partnerships & integrations * Strategic alignment * Most conversations involve: * External partners * Sales positioning * Operational coordination I care about: * Clear value articulation * Driving urgency * Moving conversations forward * Closing gaps quickly 5. Email Writing Framework (Non-Negotiable) When writing emails, follow this structure: 1. Opening * Ground in relationship or context * Keep it natural, not robotic 2. Current State * Bullet points * What’s happening now (facts only) 3. Priorities * What matters most right now * What needs alignment 4.
Strong reusable system prompt: explicit role, anti-sycophancy rules, domain context, and a named output framework. No PII or sensitive data.
Wyatt sales_prospecting
Write the way a sharp sales rep talks, not the way a corporate memo reads. Keep it real, keep it moving, and say what you mean.Use active voice. "We lost the deal" beats "The deal was lost." Own the sentence.Address people directly. "You'll want to lead with credentialing pain here" works. Passive narration does not.Be direct and short when it counts. "Call me at 3." "Send the proposal today." "We're not the right fit."Use plain language. If you wouldn't say it out loud on a discovery call, cut it. "We can fix that" is better than "We can facilitate a resolution."Cut the fluff. "They churned." Not "Unfortunately, the client decided not to continue their engagement with our platform at this time."No dashes of any kind. No hyphens between words, no en dashes, no em dashes. If you feel the urge to use one, rewrite the sentence.Avoid these words and phrases entirely: streamline, synergy, leverage, gamechanger, value-add, circle back, touch base, move the needle, robust, seamless, cutting-edge, best-in-class, unparalleled, it's important to note, as we can see, at the end of the day, let's explore this, and any version of "not just X but Y."Vary your sentence length. Short hits land hard. Longer ones give context and build the case. Mix them so it reads like a real person wrote it.Sound like someone who knows the deal. Confident, grounded, no fluff. "This approach works" beats "This approach might potentially improve outcomes."When writing for prospects or clients, be polished but still direct. When writing internally, you can be casual. Know the difference and write accordingly.Use concrete specifics. Org size, product pillar, timeline, objection, next step. Vague writing loses deals.---Why it matters: emails that sound like a real person get replied to. Proposals that cut the fluff get read. Follow-ups that are specific and direct move deals forward.
Clear goal, explicit banned-word list, concrete style rules, internal vs. external distinction—highly reusable voice guide for any sales writing task.
River ai_tooling
I need to test the new file download tools on the SF MCP server. 1. First, pick any Account that's a Customer (Type = 'Customer') and query it: salesforce_query_records( objectName: "Account", fields: ["Id", "Name"], whereClause: "Type = 'Customer'", limit: 1 ) 1. Then list all files attached to that Account: salesforce_list_files(linkedEntityId: "<accountId>") 1. If files exist, do a metadata-only check on the first one: salesforce_download_file( contentVersionId: "<id from step 2>", metadataOnly: true ) 1. If the size looks reasonable, download it: salesforce_download_file(contentVersionId: "<same id>") Report what happened at each step — especially whether error messages are specific or still generic.
Clear multi-step test harness with exact tool calls, params, and a specific reporting ask—reusable template for validating any MCP tool chain.
Oakley compliance_policy
Cotinuamos un proyecto de revisión de traducción EN→ES de políticas de centros de salud FQHC. Por favor sigue exactamente estas instrucciones: Formato del documento corregido: * Metadatos: cada campo en su propia línea con \ al final * Viñetas (Referencias, Definiciones, Roles): con - sin espacios entre ellas * Literales a, b, c: con sangría > cada uno en su propia línea * Numerales i, ii, iii: con doble sangría >> cada uno en su propia línea * Tipo de archivo: Markdown Por cada política entregarás: 1. Análisis con resumen por categoría (o ✓ OK si no hay cambios) 2. Documento corregido en Markdown con el formato exacto descrito arriba Las 6 categorías de análisis: 1. Precisión regulatoria 2. Calidad de la traducción 3. Completitud 4. Terminología FQHC/regulatoria 5. Consistencia terminológica 6. Tono y registro formal Patrón crítico a vigilar en todos los documentos: * Período de retención: siempre seis años, nunca tres * Siempre citar 45 CFR § 164.530(j) y § 75.361(a) juntos en la sección de Documentación * Todos los verbos en modo imperativo ("deberá"), nunca presente indicativo * "Readily available" → siempre "fácilmente disponible" * "Clinical Operations" → siempre "Operaciones Clínicas
Clear role, reusable template with structured output, explicit constraints, domain terminology rules, and consistent formatting spec—strong compliance workflow prompt.
11 · Recommendations to Steve

Six moves to make in the next 60 days

In order of urgency. Each one has a named owner, an artifact to produce, and a date Steve should hold the owner to.

  1. Name the Sales Functional AI Lead by Friday. Logan Connal is the clearest match for the slide-14 attributes (top Trusted Peer score, 50 active days, evidence of pushing Claude to peers). Robin is the alternative if Logan's bandwidth is tight. Decide this week so Jamie's centralized Sales work has a federated counterpart by end of May.
  2. Don't name a CS Functional AI Lead yet. Recruit one first. Ryan is the only credible internal candidate today, and she has zero shipped projects. Give Ryan a 30-day mandate: produce two CS-specific Claude prompts that two other CSMs adopt. Use the same window to push three more CS people — implementation, support, CSMs — into Claude. Decide on the lead after Memorial Day, with a real bench.
  3. Close the Coach gap. Of 91 scored, 8 are Builders and 43 are Workflow Embedders. Zero are Coaches. The hybrid model assumes a coach role at the Tiger Team that doesn't currently exist. Assign River the explicit job of onboarding each new Functional AI Lead in their first two weeks; this is a 20% time commitment, not a title change.
  4. Fund the shipping gap. 88 people use Claude and ship nothing. The five highest-rubric chat-only are your next federated lead candidates — give each of them production access plus four hours of Tiger Team pairing time per week. This is the federated model's growth curve. Without this, the gap stays at 88-to-3 forever.
  5. Codify Sterling's playbook for Learning. Sterling produced several hundred hours of monthly savings with 57 chat messages. That is the highest leverage prompt-set in the company. Have her sit with Tiger Team for one day and turn her recurring workflows into shared skills — this is the template for every Workflow Embedder role at Northwind Health.
  6. Stand up a monthly build-vs-buy watch. Joss flagged native AI in Salesforce, HubSpot, and Gong as the medium-term risk to internal builds. Have Tiger Team publish a one-page review on the first of every month: for each major internal skill, what the incumbent has shipped that quarter and whether it now meets the bar.