Agentic Code Review Systems Engineer
LangGraph forensic codebase evaluator using a Digital Courtroom protocol: evidence-gathering detectives, three judge personas, deterministic conflict resolution, and hallucination policing.
"Made repository evaluation adversarial and evidence-bound instead of letting a single judge model grade from vibes."

Automated code evaluators can hallucinate file evidence, collapse into one weak opinion, or miss security and production-readiness failures when grading complex repositories.
Designed a fan-out/fan-in LangGraph workflow where investigators collect AST, git, security, PDF, and vision evidence, judges argue independently, and Chief Justice rules apply deterministic vetoes and synthesis.
Automaton Auditor evaluates repositories through staged graph execution. It gathers objective evidence, merges it with integrity checks, asks prosecutor/defense/tech-lead personas to score independently, and writes a final audit report.
Case studies in similar engineering domains.
Event-sourced lending pipeline for document intake, extraction, credit analysis, fraud, compliance, and decision orchestration over an append-only ledger.
Multi-agent codebase cartography tool that analyzes local or GitHub repositories with Surveyor and Hydrologist agents to produce module graphs and data lineage artifacts.
Schema integrity and lineage attribution system that turns inter-system dependencies into formal contracts, detects schema/type/statistical drift, and reports downstream blast radius.