Building AI Agents · Financial Services
Building governed AI agents for financial services.
Talking about governance is one thing; building it in is another. These are working demos of AI agents for financial-services use cases — each built with the controls, scoring, and human oversight from the playbook baked in from the start. Every build ships as a live, clickable demo on synthetic data, and every one keeps a human in the loop where the decision matters.
How I build
Each project starts from a real regulated decision, defines a named metric and failure taxonomy, adds a trust-scoring and human-review layer, and maps to at least one real framework (NIST AI RMF, EU AI Act, SR 11-7). The point isn’t a clever model — it’s a model you can explain, evidence, and hand to a reviewer.
In progress
A governed agent for a financial-services decision
Pairing a working agent with a trust-scoring and human-review layer, in the spirit of a hallucination-evaluation harness but aimed at a lending, onboarding, or fraud decision. Each entry will include the problem, the approach, the governance angle, an honest note on limits, and a link to the live demo.
Live demo — check back soon.