AI finance operations command systems for regulated, capital-intensive, high-velocity organizations.
I build the control plane between finance and execution. Metrics, AI-assisted work, and the evidence behind every approval stay in one place leadership can audit.
Principles
- Show the work — public artifacts over claims.
- Synthetic data only. No employer or client information.
- Controls and auditability are features, not afterthoughts.
Proof areas
Everything here is something I can show — built systems and written arguments, not private employer work.
AI Finance Ops Command Platform
A working example of how a finance team runs AI-assisted work end to end, from the task to the review to the report leadership sees.
02Agentic Governance & Model Risk Controls
Letting AI into financial decisions without losing the approval, evidence, and escalation trail an auditor will ask for.
03Strategic Finance / Program Operations
Planning that ties capital allocation to program execution and the variance story behind every number.
Current flagship
A working platform showing how a finance team can put AI into real work without giving up the evidence trail — try it in the browser.
- Domain
- Strategic finance, program operations, governance controls
- Architecture
- Static site, synthetic data, reusable components
- Purpose
- Demonstrate the approach in public, with nothing confidential
Currently building
Active work in private repositories — published as a set once each is stable.
Fintech infrastructure
Banking data onboarding
A Plaid-based onboarding flow for pulling banking data into finance operations — the ingestion end of the control plane.
Agent tooling
Local-first document intelligence
A local-first PDF stack with a universal MCP connector, so agents can work documents without the data leaving the machine.
Operating base
Governed agentic workstation
A workstation setup for running agentic AI with the same posture as everything here: controlled, observable, QA-ready.