Generate Architecture Decision Records (ADR) & Tech-Stack Trade-off Matrices with AI
Six months from now, nobody will remember why your team chose PostgreSQL over MongoDB or RabbitMQ over Kafka. Use AI to challenge your assumptions and generate standardized Michael Nygard Architecture Decision Records (ADRs) stored right in Git.
Great architecture is about explicit trade-offs under real constraints, not chasing hype. A 1-page ADR takes 10 minutes to generate and saves hundreds of hours of future team debate.
Step-by-Step Action Checklist
0 / 4 CompletedAct as a pragmatic Principal Systems Architect who hates resume-driven over-engineering. We need to make an architectural decision on: [topic: e.g., job queue architecture / multi-tenant database isolation]. Our real-world constraints: - Team & Stack: [e.g., 3 engineers proficient in TypeScript & PostgreSQL, no dedicated DevOps] - Scale & Budget: [e.g., 200 peak RPS, <$200/mo infra budget] - Options under consideration: [Option A vs. Option B vs. Option C] Generate a Git-ready **MADR (Markdown Architecture Decision Record)**: 1) [Status, Context & Decision Drivers]; 2) [Unvarnished Trade-off Matrix]: Compare cognitive load, Day-2 operational pain, failure modes, cost and lock-in risk; 3) [Recommended Decision & Explicit Rejection Rationale]; 4) [Consequences & Re-evaluation Triggers]: What technical debt we are consciously accepting today and the exact scale threshold that will require revisiting this choice.