Labs
How the firm works. Labs is not a portfolio — it is the method that produces the portfolio: name the landscape precisely, place specific bets on where it is going, and build working systems with customers to find out which bets hold.
Two Modes of Work
Naming the landscape
Emerging fields fail first at vocabulary. Before an institution can govern an AI workforce, it needs words for what it is governing. We compress each domain we work in to its smallest set of load-bearing concepts — identity, capability, authority, trust, provenance, memory, command — so that every decision maps to something nameable.
This is slow, unglamorous work. It is also the part that compounds: a good definition outlives any product built on it.
Placing bets
Every system we build starts as a written bet: a statement about the near future specific enough to be proven wrong. Not “AI will transform banking” — that is unfalsifiable. Rather: which function goes first, what the actual bottleneck is, what institutions will pay for before they pay for anything else.
We keep the bets that fail. Knowing why a reasonable bet failed is most of what clients pay advisory firms to know.
From Observation to Advisory
The same pipeline runs behind every system in our portfolio. Here it is, with one real example traced through each stage.
Observation
Something in the field does not fit the prevailing story.
Institutions hesitate to let autonomous agents touch money, even when the agents perform well.
Definition
We compress the domain until the problem has precise words.
Trust is observable evidence of predictable behavior under uncertainty — not a feeling, a record.
Bet
We state what we believe will happen, specifically enough to be wrong.
Institutions will spend more proving agents can be trusted than making them smarter.
Working system
We build the smallest real system that tests the bet with real users.
KnowYourAgent: verified agent identity, merchant verification, immutable reputation.
Advisory
What survives becomes frameworks our clients use. What fails becomes documented judgment.
Agent governance and trust frameworks for banks preparing for agent-initiated payments.
Co-Design With Customers
None of our systems are built in isolation. Each one has design partners — banks, credit unions, and financial technology teams who bring the constraint set no lab can simulate: regulatory obligation, legacy infrastructure, and the internal politics of risk. They shape the roadmap; we test our bets against their reality.
What a design partnership looks like
Problem framing
Working sessions to define the problem in your operational terms before any system is scoped. You keep the framing whether or not we build together.
Working prototype
A real system running against your workflows — scoped narrow, instrumented for evaluation, built to answer a specific question rather than to demo well.
Shared findings
Documented results either way: what held, what broke, and what that means for your deployment decisions. Failed bets are reported as plainly as successful ones.