SHEET 01 · HOME · LIVING POLICY ARCHITECTURE
Policy compiles.Governance executes.
Dartmouth Advisory Partners builds the compiler that turns institutional policy into executable governance, present at the moment a decision is made, reviewed before any change ships, sealed to a tamper-evident record. Built for institutions that require their rulebook to be live, not static.
↓ Who this is for
Built for asset managers, OCIOs, pensions, insurers, and family offices. In Mercer's 2026 survey, 55 percent of asset managers already run AI inside at least one investment process. Grant Thornton found 78 percent of firms lack confidence they could pass an independent AI governance audit within 90 days. The distance between those numbers is operational risk. Closing it is the work.
↓ The drawing set
02 · LIVING POLICY
The four commitments.
Policy as source code, dry-run before deployment, sealed to an audit chain, revocable inside a cooling-off window.
03 · HOW IT WORKS
The compiler pipeline, live.
Watch a policy change travel from extract through compile, shadow, approve, and reconfigure, with every gate and audit step surfaced.
04 · RUNTIMES
Three runtimes, one envelope.
Overture for allocation, Falcon for reporting, Resolve for settlement, each bounded by the same compiled rulebook.
↓ Research
Open standard · June 2026
AA-1, the Agent Governance Attestation
An open standard for disclosing how AI agents are governed: which controls an organisation runs, at which tier, with what evidence. We proposed it, filed the first attestation on our own systems, and steward it in the open.
See the standard
Working paper · June 2026
From frameworks to filings: the case for a standard disclosure layer in AI agent governance
Why comparability and demand, not technical merit alone, decide whether a standard is adopted, drawn from the adoption histories of IFRS, ISO 20022, and SOC 2.
Read the paper