IBC2026 · SEPTEMBER 11-14, 2026

See UX Engine 11 at IBC

An agent that knows what an editorial list is

IBC2026 · HALL 5, STAND 5.G50 · 11 TO 14 SEPTEMBER

Your personalization engine already knows your catalog. Now agents can use it.

At IBC2026 we show what happens when LLM agents sit on top of a recommender that knows your catalog, your viewers and your editorial rules. Live demos and one benchmark that explains why we built them this way, and engineers at the stand for four days.

THEY RELY ON UX ENGINE

WHERE TO FIND US

Look for us in Hall 5

Find us at stand 5.G50, near the Content Everywhere Hub, the place where the broadcast and broadband sectors meet, serving as the central destination for technologies focusing on OTT, streaming, multi-platform delivery, and content monetization.

UX ENGINE 11 · THE AGENTIC RELEASE

Put AI agents to work with ContentWise UX Engine 11

UX Agent Tasks

Your editorial team rebuilds the same rails every week, and the rules behind them never change.

Editors delegate curation to an scheduled agent: “Keep this rail filled with available titles featuring trending actors.” “Add sports events airing this weekend, ranked by popularity.” “Drop anything expiring in seven days.”

The agent queries the web and the catalog. The editor approves.

MCP Server for UX Engine

Every agent you build needs the same four things from your platform, and building each integration by hand is how a strategy becomes a backlog.

Recommendations, search, similar items and viewer preferences, exposed as tools any AI agent can call.

Same recommender, any agent, one interface.

New Catalog Inspector

An agent is only as useful as what it knows. Thanks to the newly rebuilt Catalog Inspector, before an agent touches a rail, it is able to directly source the item metadata: genres, cast, keywords, connector IDs, all the arrays your ingest actually produced.

Catalog Inspector shows what UX Engine knows about any title, field by field.

OUR POINT OF VIEW

Generative AI in recommendations: the illusion of progress

Paolo Cremonesi, our co-founder and CTO, and professor at Politecnico di Milano, benchmarked more than 50 LLMs across four deployment scenarios. Around 90% of them fail under production conditions.

That result is why every demo below is built the way it is. The agent does not generate recommendations. It calls a recommender that has been in production for twenty years, and it works with what comes back.

BEFORE YOU COME

99 personalization use cases, and the KPIs they impact

Renato Bonomini, our VP Sales Engineering, mapped every discovery, engagement, diversity, monetization and operational scenario UX Engine supports.

Bring your platform to the stand and we will show you how many of the 99 apply to you, and which to run first.

Related resources

Agent Engine

Move beyond basic recommendations. Unify your engagement strategy with our next-gen personalization engine.

Metadata Lab

Turn your catalog into a discovery engine. Automate and enrich your metadata for better AI performance.

UX Engine

Watch viewers grow with your content strategy. Explore our flagship personalization and discovery solution.