Write the RAG-versus-fine-tuning ADR
Commit to writing an ADR for a real RAG-versus-fine-tuning architecture decision.
checklist Document a real RAG-versus-fine-tuning architecture decision A current or upcoming AI assistant where the team must decide between RAG, fine-tuning, prompt-only, or hybrid. ADR scaffold: Context: what user workflow and risk are we solving? Criteria: freshness, authority, behavior, cost, safety, eval readiness. Options considered: prompt-only, RAG, fine-tuning, hybrid. Decision: selected option and why. Consequences: what this makes easier, harder, cheaper, riskier. Revisit trigger: exact eval or data condition that would reopen the decision. In 3 days, review whether the ADR exists and whether the revisit trigger is measurable. A benefits or policy assistant where source freshness and citations are…
Sign up free — one personalized lesson every day, matched to your role and goals.
Already have an account? Sign in