Box AI Extract pulls structured fields from documents, and @OpenAI Decisions API routes them to the right workflow.
In this demo, a COI arrives in Box through a File Request. Box AI extracts coverage fields from both the upload and the project requirements. The Decisions API and returns a typed route: request endorsement, specialist review, or ready for human review.
A Box webhook triggers the whole process on document upload. The app creates a review task with context, adds the right collaborator, and writes extracted values and decision confidence back to the file as metadata.
Box is where the documents arrive and where people review them.
Watch👇
In this demo, a COI arrives in Box through a File Request. Box AI extracts coverage fields from both the upload and the project requirements. The Decisions API and returns a typed route: request endorsement, specialist review, or ready for human review.
A Box webhook triggers the whole process on document upload. The app creates a review task with context, adds the right collaborator, and writes extracted values and decision confidence back to the file as metadata.
Box is where the documents arrive and where people review them.
Watch👇
Oct 15, 2026 · 6:00 PM UTC
4
3
15
179,801
Automating enterprise workflows is not just about extracting fields from documents and routing them to the right process.
The real challenge is making every decision traceable.
When an AI system reviews a document against project requirements, teams need to know:
→ Which source fields supported the decision?
→ Which requirements were matched or violated?
→ What happens when sources conflict or requirements change?
→ Can a human reviewer verify the reasoning before approving the outcome?
Confidence scores and automated routing improve efficiency. But without traceable evidence, automation can make incorrect decisions faster.
This is why enterprise AI needs a reliable knowledge layer beneath its workflows: structured information, source-level evidence, relationships, and traceability.
This is also the direction we’re exploring with @ekosproject , an open-source enterprise knowledge compiler that turns source data into structured, evidence-backed knowledge for AI agents.
📖 Related article — EKOS: From Enterprise Data to an Evidence-Backed Knowledge Layer for AI Agents
medium.com/@Omid.EKOS/eko…
🔎 Explore the EKOS Web Console demo
alexeyban.github.io/EKOS/presentat…
💻 Open-source project
github.com/alexeyban/EKOS
The goal isn’t just to automate decisions.
It’s to make those decisions traceable, verifiable, and auditable.
@ekosproject @Box @levie
#EnterpriseAI #AIInfrastructure #AgenticAI #OpenSource
0
0
0
43