AI & Document-Automation Engineer

AgentA — self-improving engineering harness

An autonomous system that improves its own code behind a 7-tier verification gauntlet — parse, unit, property, mutation, benchmark, sealed holdout, fresh-context critic — so improvements are earned, not hallucinated.

Improvements gated by a sealed holdout + fresh-context critic

Inner research loop ported from Udit Goenka’s autoresearch (MIT, after Karpathy); the meta-improvement architecture and verification stack are my own.

Financial-Statement Extractor

Turns a government audit statement (PDF) into clean structured JSON, then re-derives every total to prove the extraction is correct, with a golden-file regression test. The model never does the arithmetic — tested code does.

25 checks, 0 exceptions — inject one wrong figure and it’s caught

GrantLedger

Full-stack B2B SaaS that auto-categorizes nonprofit grant spending into 2 CFR 200 budget categories, tracks budget-to-actual per grant, and generates audit-ready compliance reports, with QuickBooks/Xero integration.

Audit-ready compliance reports, budget-to-actual per grant

NeuroFour — Connect 4 strength-per-byte

Connect 4 is a solved game, so perfect play is a known, fixed target. NeuroFour ranks 19 Connect-4 agents by strength per byte under a hard 5M-FLOP-per-move budget, scored against an exact solver.

Zero leads optimality, ladder Elo, and NeuroFour Score among the 19 under-budget agents, at 0 bytes, ahead of all 14 learned-net entries (6 distinct weight files, largest ~24 KB): ladder Elo 754 vs 631 and NeuroFour Score 96.45 vs 74.25 — a score that divides strength by a size penalty — optimality 0.960 vs 0.957 by a single move in 300. Each “vs” is the strongest learned entry on that axis (optimality is a tie); the four entries behind them load just two weight files. The cost: Zero spends 4,999,028 of the board’s 5M-FLOP budget — 99.98% of it, 18th of 19 on FLOPs/move (and 4th on soundness, 15th on latency).

Free-tier API — first move may take ~30–60s to wake.