Resources
Insights on building AI that runs real operations
White papers, field notes, and frameworks from the team building the governed layer for enterprise operations.
Watch
ProductSep 2026
Building an Invoice Audit Agent From an Empty Workspace
An unedited twenty-five-minute build. A freight invoice audit and dispute workflow, described in plain language and assembled into a governed operating model, deployed agents, and a first run that escalates to a human for the right reason.
Read the walkthroughFeatured writing
AI StrategyAug 2026
The Sand Castle Is the Strategy
AI has made the prototype nearly free: a working model of a business process in a day, a convincing demo in a week. Most enterprises then make one of two mistakes. They promote the demo to production, or they throw it away and pay for the discovery twice. The third path is to treat every prototype as a survey of the ground, and to keep what it finds.
Read the essay
AI GovernanceAug 2026
Intelligence Sovereignty: Owning Your Alpha in a World of Rented Minds
Alex Karp says enterprises are handing frontier labs the weights and alpha of their businesses. He's right, and the fix isn't training your own model. It's an ontology at the core, a governed runtime around it, and a signed receipt on every turn. We're building exactly that, and it runs today.
Read the paper
Industry AnalysisJul 2026
The Model Is Not the Product: Why Reliable AI Runs on an Ontology
A short argument and a surprising result. The usual fix for unreliable AI agents is a bigger model. We think it is the layer around the model, and we ran the experiment to prove it. The full data and methodology are in the Technical Report.
Read the paper
Industry AnalysisJul 2026
The Sisterhood of the Travelling Ontology: Governing Agents Across Every Runtime, On Infrastructure You Never See
We already showed that reliability lives in the layer around the model, not the model itself. The same fact has three larger consequences: the runtime becomes swappable, the data boundary moves to the customer, and the enforcement can prove itself. Governance turns out to be a property of the ontology.
Read the paper
Industry AnalysisApr 2026
The Cost of Intelligence Has Collapsed. The Cost of Execution Is at an All-Time High.
Custom AI demos in 15 minutes. Models too capable to release. The gap between what's possible and what's safe to deploy has never been wider, and it's reshaping how enterprises should think about technology partnerships.
Read the paper
EngineeringJan 2026
Why We're Building OrgBench™: You Don't Need AGI to Get AGI-Level ROI
High IQ doesn't predict success. Neither does raw model intelligence. The real bottleneck is orchestration, and that's why we're building OrgBench™.
Read the paper
EngineeringDec 2025
Harness Engineering: How We Make AI Reliable for Supply Chain Operations
Why smarter AI models alone won't solve reliability problems, and how we turn probabilistic AI into deterministic performance.
Read the paperAll articles