Barndoor acquires Diaphora to govern AI workflows for enterprises
What's the deal? Barndoor AI, which describes itself as an AI Gateway for enterprises, has acquired DiaphoraDealroom has a profile for this one. Try Dealroom →, the startup behind Frags, an open-source engine for building AI workflows. The deal, announced September 16, 2026, is an acquisition; terms were undisclosed.
The combined product lets enterprises build Blueprints — repeatable AI workflows that connect tools and data through defined steps that can be governed and distributed across authorized teams. Barndoor pairs Diaphora's workflow technology with its own governance and access controls.
Why now? Enterprises have rushed to adopt AI tools, but turning experiments into repeatable workflows remains hard. Barndoor points to three barriers: reliability, access, and distribution.
The team: The full Diaphora team joins Barndoor. The startup began as an independent project by Simone Pezzano, with Jay Parisi collaborating as the technology evolved; Barndoor chief executive officer Oren Michels later became an advisor. The companies call the deal a "spin-in."
The product: Diaphora is built on Frags, an open-source runtime that uses the Frags Modeling Language to define how AI executes a workflow and constrain where a large language model can act. The engine will remain open source, letting anyone contribute, extend, or verify it. Within Barndoor, automations become governed MCP tools, with role-based controls deciding who can discover and run each one.
A Blueprint connects tools and data through defined steps to create a repeatable workflow, using AI only where judgment is required. If it cannot complete a required step, it stops and flags the issue rather than filling the gap.
What's the endgame? The goal is to bring AI automation to higher-stakes enterprise work that still needs manual execution. "Most AI systems operate from a plan, what you hope is going to happen. A blueprint defines what actually happens," said Michels.
The signal: The deal reflects a market shift from AI pilots toward governed, repeatable deployment — pairing predictable execution with the access controls enterprises need to scale automation safely.
Image credit: Generated with Gemini
