Your browser is out-of-date!

Update your browser to view this website correctly. Update my browser now

×

Unlocking AI in production

Florian Reimann, CEO of Yamdu, discusses the purpose of IBC Accelerator, FRAMES: Federated Retrieval, Agentic Media Environment and Software (Defined Workflows)

FRAMES aims to prove an end-to-end intelligent media production framework that dissolves traditional silos by standardising archival asset management and its metadata and enforcing Zero Trust security.

A small Yamdu team is working closely with the MovieLabs ontology group—we’re already active in the MovieLabs Industry Forum, so the OMC (Ontology for Media Creation) alignment work didn’t start with this project. However, Yamdu’s API has been heavily extended to fulfil the requirements for this software-defined workflow.

Why do you believe the project is important for the media industry?

Everyone talks about AI in production, but agents are only as good as the structured context you can hand them. Today most production data is trapped in documents built for humans to read. Making it machine-readable through a shared ontology is the unglamorous prerequisite for everything else—interoperability between vendors, automation that doesn’t require a human to retype the same information five times, and AI that can answer questions about a production instead of guessing.

A common language also means smaller vendors can plug into studio pipelines without bespoke integration work, which matters for the European market in particular. We are proud that Yamdu can play a small but important role by streamlining production data and the content of a script from day one in order to be presented in a structured way, which is useful for AI–and downstream tools and vendors.

REGISTER FOR IBC2026

What’s been the biggest challenge you’ve faced as part of the project?

Mapping messy reality onto a clean ontology. Real productions don’t behave like data models—a scene gets split, a character is credited three different ways, a location is both a set and a real address. Deciding where Yamdu’s internal model bends to OMC and where OMC has to accommodate how productions actually work has been the hard part, and it’s a conversation, not a one-off mapping exercise.

Why did you want to be part of this year’s IBC Accelerator programme?

Because standards work only proves itself when something is running end to end with other people’s systems. The Accelerator forces that: a fixed deadline, real partners, and a demo that either works or doesn’t. It’s a faster and honest test of OMC and Yamdu’s new third-party API with real-time-feedback loops.

What can IBC Show attendees expect to see from you in Amsterdam?

Alongside MovieLabs, the Yamdu team is looking forward to live demo in the Accelerator Zone — a real production dataset served over our OMC-aligned API, and agents querying it to do useful pre-production and production work. We are happy to show why this matters for studios and for small production companies alike.

See all of this year’s projects at the IBC Accelerator Zone in Hall 14.