Production AI · Model Evaluation & Adaptation · Payments · Risk · Financial Operations
I build systems at the intersection of AI, financial operations and technical delivery.
My work turns complex operational problems — payment reconciliation, failed-payment recovery, credit/risk workflows, financial visibility and AI evaluation — into measurable, auditable systems.
I work across the full delivery path:
Discovery → Architecture → Data → APIs → Business Logic → AI/ML → Evaluation → Deployment → Monitoring
I’m building IntelligenceOS in public: an applied AI system spanning domain-data preparation, evaluation, model adaptation, agent workflows, verification and production deployment.
Current work includes:
- ML and LLM evaluation infrastructure
- Finance/risk dataset engineering
- Model adaptation and fine-tuning experiments
- Production AI deployment patterns
- Financial operations systems
| Project | Focus | Evidence |
|---|---|---|
| eval-harness | Model evaluation & financial-risk ML | Reproducible benchmark |
| ios-risk-data-foundry | Domain data engineering | Validated pipeline + benchmark |
| multi-processor-reconciliation | Payment reconciliation architecture | Reference implementation |
| payment_recovery_engine | Failed-payment recovery | Validated simulation / reference implementation |
| cashflow-forecasting-engine | Cash forecasting & decision support | Reference implementation |
Python · PostgreSQL · Supabase · APIs · Financial Systems · AI Evaluation · RAG · Model Adaptation · Workflow Orchestration · System Design · Testing · Deployment · Observability
Public projects are explicitly labelled as:
Production · Anonymized Production Case · Validated Prototype · Reference Implementation · Benchmark
Simulated, benchmarked and projected results are never presented as verified client outcomes.
Building IntelligenceOS and documenting the engineering behind production AI + financial systems.