Current focus.

My public engineering portfolio centers on AegisGraph, CreditOps AI, TradeFlow, and CreditFlow QE. AegisGraph presents deterministic security investigation, CreditOps a recorded synthetic trading-incident showcase with public source, and TradeFlow reproducible trading-workflow tests with a Railway-hosted synthetic demo behind Cloudflare Tunnel. CreditFlow QE adds a public synthetic banking simulator with Java API/browser automation and separately recorded hosted and local validation.

The supporting security work covers CI evidence gates, detection tuning, synthetic cloud reassessment, and matching-rescan closure. Each case study distinguishes its fixture or recorded evidence from production claims. The portfolio and all four flagship demos now run on Railway. Current case studies document TLS, database authority, off-volume backups and source-release boundaries, while the AegisGraph article preserves its original design and dated updates. The October 25 TradeFlow compatible dependency patch is deployed, with a dated validation record and blocking full/production npm audits.

Broader live-model evaluation and authenticated tenant controls remain outside AegisGraph's public demo. CreditOps recordings leave initiating causes unconfirmed. TradeFlow's tests establish finite synthetic behavior without exchange execution; CreditFlow QE similarly has no real card or funds movement or banking SLA. CreditFlow's first scheduled daily-backup closeout and retained-resource cleanup remain pending as of October 25. The lab notebook keeps open questions tied to public evidence.