6.1 · Run the A/B test — vanilla vs. ontology
The starter ships a measured proof:validation/ab-test-vanilla-vs-ontology.md — seven questions asked in two threads against the same connector, one with the ontology repo connected and one without, scored on correct / consistent / traceable. The governed answers are pre-filled from validated golden values.
Prompt
You’ll see: the coded, multi-definition, and risk questions land exactly on the governed values, while the control question (“how many members total?”) shows it’s not magic — it’s governance where governance matters. The headline: correct on the hard ones, consistent across phrasings, traceable every time.
6.2 · Run the golden queries
The starter’s golden values are already pinned and verified against the demo dataset (seevalidation/golden-queries.md and the end-to-end script in validation/run-golden-queries.md). Against your warehouse, re-pin them to numbers you trust:
Prompt
You’ll see: accuracy checked, not asserted — and a triage of any mismatch into data vs. definition vs. window.
6.3 · Customize a definition
Your organization inevitably defines something differently — a readmission exclusion, a line-of-business filter, a different risk-model year.Prompt
You’ll see: the change land as a reviewable PR in your repo — the template stays pristine upstream; your adaptations are yours.
6.4 · Localize the vocabulary
ontology/notes/glossary.md holds the canonical claims terms — member, member-month, claim, encounter, charge vs allowed vs paid, denial — each with a line-of-business variance column flagging where commercial, Medicare, and Medicaid deployments diverge (retroactive Medicaid eligibility, MA encounter completeness, HHS-HCC/CDPS instead of CMS-HCC…).
Prompt
You’ll see: the vocabulary localized in one reviewable pass — so “member month” and “allowed amount” mean your org’s thing, everywhere, from now on.
Two habits as you make it yours —1 · Write for the search box. As you extend the kit, keep a short README per folder and repeat the phrases your teams actually use (metric names, synonyms, team names) in the prose — future threads find context by search, not browsing.
2 · Let usage drive the roadmap. Stand up a weekly gap-review playbook: mine repeated questions, manual SQL, and mid-thread corrections; have Ana draft small reviewable patches; a named owner approves. The kit is the seed — usage is what grows it. (See Ontology Operations Module 4.)
2 · Let usage drive the roadmap. Stand up a weekly gap-review playbook: mine repeated questions, manual SQL, and mid-thread corrections; have Ana draft small reviewable patches; a named owner approves. The kit is the seed — usage is what grows it. (See Ontology Operations Module 4.)
✅ Checkpoint
- Golden queries ran; any drift was triaged (data / definition / window)
- One definition is now yours — PR’d, noted, and pinned with a test