Writing for the finance teams designing this from scratch.
Cornerstone pieces on building AI agents that auditors will sign off on. No vendor copy. Citable code, real audit shapes, and the trade-offs we hit.
Cornerstone articles
Why Fixed-Income Desks Need FINOS AI Governance Now
The cross-framework alignment of FINOS AIGF v2.0 with NIST AI RMF, EU AI Act, ISO 42001, and SR 11-7 — and why ad-hoc AI policies will not survive a 2026 audit.
Materiality-Driven Access Control: Why Tier Routing Matters for Bond Desks
T0 to T3 tier routing for AI agents on a fixed-income desk — with a worked $5M corporate-bond example and the split-trade bypass nobody talks about.
The Four Pillars of Governed AI in Finance
Policy gate, audit chain, HITL envelopes, and the AIGF coverage gate — the four primitives every agentic-AI system in a regulated workflow needs.
How to Build an AI Agent Your Auditor Will Actually Approve
Hash-chained audit logs, framework_ref schemas, pure-function policy gates — and the three things a real auditor checks first.
Field notes
Open-Source vs Vendor SaaS: Building Your Own Governed AI Desk
Why self-hosting an AI agent for the buy-side desk is a board-level conversation in 2026 — and the fork-and-swap pattern that makes it tractable.
FIX Gateway as AI: Connecting Your OMS to Claude-Backed Quant Analysis
How BondFoundry's FIX 4.4 gateway sits in front of an AI agent — with HITL approval envelopes, ExecutionReport flowback, and a HITL queue that doesn't trust the model.
Evaluating Agentic AI: The Four-Dimension Battery
Accuracy, policy, robustness, latency — the four eval dimensions every agentic-AI system in finance should be gated on, with a CI gate at 85% AIGF coverage.
From CSV to Real-Time: Curve Provider Architecture for Multi-Source Bond Markets
How BondFoundry ingests yield curves from ECB, US Treasury, Bloomberg BVAL, Refinitiv RDP, and CSV — with stale-curve detection that survives the morning.