By the end of this lesson you will be able to run a kickoff, find the four key people, and write a one-page engagement brief.
By the end of this lesson you will be able to run a kickoff, find the four key people, and write a one-page engagement brief.
ENGAGEMENT BRIEF — Brightline Health — v1 Problem: Referral intake takes ~3 days: faxes hand-keyed into the EHR by 6 people. Metric: ??? discovery gap Scope: Fax OCR + field extraction + review queue. Not scope: EHR write-back (phase 2). Sponsor: ??? discovery gap Champion: Priya Nair (Intake Lead) Gatekeeper: Sam Okafor (IT Director), security review BAA: ??? discovery gap Asks: Sandbox credentials (Sam, Mar 8). 2 hrs/wk of Priya.
Top risk: No named sponsor: a loved pilot that never expands. Make this your top priority before Friday.
Monday is your Brightline kickoff with Dana Ruiz. By Friday she'll have decided whether you're a vendor to manage or a partner to trust.
Beforehand, read the sales notes and contract and write down what was promised and your questions. Sales notes are a starting hypothesis, not a spec: they were written to win the deal, so they tend to round up what the product can do and round down how messy the data is.
Leave the kickoff with the problem and success metric, the people, early risks (data access, security, competing projects, budget cycles), and actions with an owner and a date. If an action has no owner, it won't happen; if it has no date, it will happen late.
| Role | Who | What they need from you |
|---|---|---|
| Executive sponsor | Owns budget and business outcome | Brief, regular proof the investment works |
| Champion | Day-to-day advocate, often a team lead | Quick wins to show their boss |
| Users | People whose workflow changes | Time saved, and being listened to |
| Gatekeepers | IT, security, legal, data owners | Clear documentation, early involvement |
At Brightline: Dana is the sponsor (she holds the pilot budget), Priya Nair, the Intake Lead, is the champion, the six intake staff are the users, and Sam Okafor, the IT Director, is the main gatekeeper.
The most common way engagements die: a champion but no sponsor. The pilot is loved but never expands. No named sponsor by Friday? Make it your top priority.
Meet gatekeepers early with a data-flow diagram and ask what they need to approve. Gatekeepers rarely block projects out of spite. They block projects that surprise them. A one-page diagram showing where data goes, what's stored and who can see it turns a three-week security review into a one-week one.
By the end of week one, send the sponsor and champion one page:
ENGAGEMENT BRIEF — Brightline Health — v1 (March 6)
Problem: Referral intake takes ~3 days because faxed referrals are
hand-keyed into the EHR by a team of 6.
Goal: Same-day intake for 80% of referrals by June 30.
Metric: Median hours from fax received → referral in EHR.
Baseline: 71 hours (Feb sample, n=412).
Scope: Fax OCR + field extraction + human review queue.
Not scope: EHR write-back (phase 2, pending vendor API access).
People: Sponsor: Dana Ruiz (VP Ops) · Champion: Priya Nair (Intake Lead)
Security review: Sam Okafor (IT Director)
Risks: EHR API approval may take 6+ weeks. Handwritten referrals ~15%.
Referrals are PHI: no real faxes to the model until a BAA
is signed and HIPAA scope is confirmed (Sam + account team).
Next 2 wks: Profile 1 month of faxes on-site; extraction prototype on 50
de-identified samples until the BAA is in place.
Asks: Sandbox credentials (Sam, Mar 8). 2 hrs/wk of Priya's time.
A line you can't fill is a discovery gap. Once the sponsor replies "looks right", it's your written scope for week-5 change requests. Try the diagram above: mark lines as gaps and watch the risk line change. A missing sponsor always ranks first.
Send it, with the gap visible: "Sponsor: to confirm. Who signs off on expanding beyond the pilot?" A brief with an honest gap is useful; waiting for a perfect one isn't. Then make finding the sponsor your top action. At Brightline, Dana holds the pilot budget, but expansion needs the CFO, who wants a measured result. That's exactly the kind of fact the brief should surface in week one.
Brightline's referrals are protected health information (PHI) under HIPAA, and Brightline is a covered entity. That changes the order of work:
Put the BAA on the risk line of your brief with an owner and a date. In regulated work it is often the critical path, ahead of any code.
Not "yes." Say that nothing with PHI goes to any vendor until a BAA is signed; that Anthropic offers HIPAA readiness with a BAA for eligible Claude API features, but not for every platform or feature, so you'll confirm the exact scope in writing with the account team; and that until then you'll prototype on de-identified or synthetic samples and log only request IDs, timings and token counts.
[ ] Notes shared within 24 hours; the four roles named
[ ] Metric agreed with a baseline (or a plan for one)
[ ] Data access requested; security review started; for regulated data,
the BAA owner and timeline known
[ ] Brief v1 acknowledged; weekly update agreed
Cadence: a short written update, same day every week: Shipped, Metric, Next, Need from you. Keep it to what fits on a phone screen. The sponsor reads it between meetings, and the consistency matters more than the length: a missed week reads as a problem even when there isn't one.
Cheat card
- Leave the kickoff with a metric, the four named roles (sponsor, champion, users, gatekeepers), risks and actions with an owner and a date.
- No sponsor by the end of week one is your top risk: champion-only pilots are loved and never expand.
- With PHI, the BAA and confirmed HIPAA scope come before any real data reaches a model; prototype on de-identified data and log only IDs, timings and token counts.
Problem / Goal / Metric + baseline / Scope / Not scope / People / Risks / Next 2 wks / Asks Weekly: Shipped · Metric · Next · Need from youInterview line "By Friday of week one I want a one-page brief the sponsor has replied 'looks right' to. Every line I can't fill is my discovery list for week two, and a missing sponsor goes to the top."