Enterprise AI governance provider Barndoor has acquired Diaphora, a startup developing technology for building repeatable AI workflows, as the New York City-based company hopes to combine automation development with controls over AI access and execution. Given recent news reports of AI agents acting without the knowledge or support of their creators, governance is becoming a far bigger issue for enterprises employing agentic systems.
Diaphora, also headquartered in New York, developed Frags, an open-source workflow engine that uses the Frags Modeling Language to define how AI executes workflows and limit where large language models can act. Diaphora’s team will join Barndoor, while Frags and its modeling language will remain open source, the companies said.
The combined technology will enable organizations to create what Barndoor calls Blueprints—AI workflows that connect tools and data through predefined steps. AI can be incorporated into portions requiring judgment, while other steps execute according to the workflow.
“Most AI systems operate from a plan, what you hope is going to happen. A blueprint defines what actually happens,” says Oren Michels, co-founder and CEO of Barndoor. “By combining Diaphora’s workflow technology with Barndoor’s governance, enterprises can make AI automation more predictable, easier to deploy and accessible to employees without requiring them to become experts at directing AI.”
Barndoor provides role-based controls governing which tools, models and data AI systems can access. Diaphora workflows will integrate those controls, enabling organizations to determine which employees can discover and execute individual automations.
Financial details of the acquisition were not disclosed.