Search papers, labs, and topics across Lattice.
This paper introduces RADAR, a multi-stage automated code review system deployed at Meta, designed to address the growing gap between AI-generated code and reviewer bandwidth. RADAR classifies diffs by authorship and source, applies eligibility gates, static heuristics, a learned Diff Risk Score, LLM-based review, and deterministic validation before landing changes. Results from reviewing 535K+ diffs show RADAR significantly reduces revert rates (1/3 of non-RADAR) and production incidents (1/50 of non-RADAR), while also decreasing median time to close by over 330% and median diff review wall time by 35%.
Meta's RADAR system proves you can automate code review at scale for AI-generated code, slashing review times by 330% while *also* dramatically reducing revert and production incident rates.
AI-assisted coding tools have altered software production. At Meta, significant lines of code per human-landed diff grew by 105.9% year over year and per-developer diff volume rose 51%, with agentic AI responsible for over 80% of that growth. Meanwhile, the share of diffs receiving timely review has declined, exposing a widening gap between code supply and reviewer bandwidth. We ask three questions that progress from feasibility through calibration to impact: (1) can risk-stratified automation operate at scale across diverse organizations, (2) how does tuning the risk threshold affect the trade-off between automation yield and safety, and (3) to what extent does automated review reduce end-to-end latency for AI-generated changes? We deployed RADAR (Risk Aware Diff Auto Review), a multi-stage funnel that classifies each diff by authorship and source type, applies eligibility gates, static heuristics, a machine-learned Diff Risk Score, LLM-based Automated Code Review, and deterministic validation before landing qualifying changes. We evaluate RADAR through telemetry covering 535K+ RADAR-reviewed diffs, observational before-after comparisons for policy changes, and difference-in-differences analysis of efficiency outcomes. RADAR has reviewed 535K+ diffs and landed 331K+. Relaxing the Diff Risk Score threshold from the 25th to the 50th percentile increased the approve rate to 60.31%. The revert rate for RADAR-reviewed diffs is 1/3 that of non-RADAR diffs, and the Production Incident rate is 1/50 that of non-RADAR diffs. RADAR reduces median time to close by over 330% and median diff review wall time by 35%. Risk-aware layered automation can materially reduce review bottlenecks created by AI-driven code growth without compromising production safety.
