FSI technology executive · 25 years · AI-native builder.
I've spent 25 years where financial services meets technology: rescuing a troubled wealth-platform program and reframing it into a $750M scoped multi-year commitment, bringing a ransomwared business back with no DR (MVP in about 50 days, full Azure migration in 75), and growing a client portfolio from $7M to $140M. Now I build the governance that lets autonomous AI operate inside institutions that answer to examiners, risk committees, and 3am incidents.
One framework, A0→A4 deployment authority, across six co-equal regulated verticals, each encoded as runnable, DOI-archived Python. A non-overridable sovereign veto, a hash-chain audit ledger, hard envelopes, and mechanical demotion: the controls that turn "the agent decides" into "the agent decides, and you can prove what it was allowed to."
Framework + whitepaper → autonomy-ladder.io · The six-library family → autonomy-ladder-libraries
| Library | Vertical | Primary rule | Tests |
|---|---|---|---|
| finserv-agent-audit | Cross-vertical financial services | EU AI Act Art. 14 · SR-letters · FCRA | 722 |
| banking-agent-audit | Banking | ECOA/Reg B adverse action · BSA/AML/OFAC | 182 |
| payments-agent-audit | Payments | OFAC · Reg E · rail finality | 183 |
| payer-agent-audit | Health-insurance payer | NAIC Model Bulletin (UM / prior auth) | 156 |
| private-capital-agent-audit | SEC-registered advisers | Advisers Act §206 | 181 |
| cre-agent-audit | Commercial real estate | Fair-housing · FCRA §607(b) | 348 |
All six: zero runtime dependencies · mypy --strict · SHA-pinned CI · golden corpora of real, primary-sourced enforcement actions · DOI-archived for citation.
I run NTCI as a small, self-funded research lab. Its work tests the same governance ideas the public libraries encode: what an agent may do on its own, when a person has to sign off, and how you prove afterward what happened. Hundreds of build sessions, a running mistakes log and a long record of architecture decisions have taught me a lot about running frontier models under real constraints. The public libraries are the part anyone can inspect.
$7M → $140M portfolio Cognizant 3× Partner of the Year (top-five US bank)
$12M → $40M turnaround Wipro Opus 12-day ransomware · COVID · Azure in 75 days · track to SOC 2 + ISO 27001
$750M commitment Broadridge wealth-platform delivery rescue
Building NTCI AI-governance reference libraries · autonomous-systems research
- 📍 Dallas–Fort Worth, TX
- 🎯 CTO · CPO · Chief AI Officer · Field CTO in FSI, fintech and private capital
- 🤝 Full-time, fractional and advisory
- ✍️ Writing on AI governance in regulated industries, autonomous systems, and crisis leadership
"The long game took 13 years to compound."