We need a Python backend engineer toproductionize and harden an existing vehicle analytics platform. This is NOT a greenfield build — the system is functional, but parts are stubbed, random/non-deterministic, or missing production safeguards. Core goals Replace placeholder logic with production implementations Ensure deterministic outputs for pricing + similarity ranking (no randomness) Make WebSockets reliable (auth + per-user delivery + multi-instance safe via Redis) Implement AI smart review endpoint with web evidence + caching/versioning Fix SQL bugs, add constraints/indexes, improve query performance Add proper pytest + async test coverage (CI-ready) Tech stack FastAPI • SQLAlchemy • Redis (async preferred) • WebSockets • Relational DB • LLM API integration • Pytest/pytest-asyncio Deliverables Stable WebSocket notification system (authenticated, multi-instance safe, per-user delivery) Deterministic pricing benchmark engine + backfill for existing data Deterministic similar listings engine (weighted scoring + tie-breaking + widening fallback) AI smart review endpoint (evidence-backed, cached, versioned, safe fallbacks) DB constraints/indexes + migrations notes Tests covering critical flows, runnable in CI Engagement Contract, milestone-driven. When you reply, include Estimated total hours Breakdown by workstream (WebSockets / pricing / similarity / AI review / DB / tests) Risks/unknowns you foresee Proposed milestone plan Testing strategy Required access (repos, staging, DB/Redis, API keys) Anti-spam: Start your message with “BENCHMARK” and confirm you reviewed the attached technical hiring brief.