AI/ML Architect — 1–2 Hr Feasibility Review of a Contract-Review AI Agent Plan

We're building an AI agent that ingests commercial contracts, extracts key clauses and obligations, checks them against our playbook, flags risk, and answers portfolio questions with citations. We have the engineering team and a detailed plan (attached) — but no senior AI/ML architect. We need an experienced architect to spend 1–2 hours reviewing the plan and giving a candid feasibility assessment. What we need from you: - Sanity-check the architecture, tech stack, and the API-first vs self-hosted hosting decision - Pressure-test our accuracy/grounding strategy and the ≥90% extraction-F1 target for legal use - Tell us whether the 18-week timeline and team are realistic - Flag the top risks we've under-weighted, and give a go / no-go / go-with-changes call Ideal background: hands-on shipping LLM/RAG or agentic systems in production, ideally with document-heavy or regulated/legal data, plus a real opinion on build-vs-buy. Deliverable: a short written review (1–2 pages, or comments annotated on our doc). The attached brief has full requirements, architecture, hardware options, timeline, and the specific questions we want answered. To apply, start your proposal with the single biggest risk you see in a plan like this.