PRODUCT
AI.
Layers of intelligence.
AI in Doors works on multiple fronts: inside the EMS to empower your teams, and in your applications to create richer, smarter experiences for audiences.
AI inside the EMS.
Doors includes an MCP server that exposes a set of tools to create resource events — the atomic changes that drive everything in the EMS.
Our MCP client connects to this server and acts as an agentic assistant. It understands natural language and can:
→
Create or modify layouts, workflows, logics, timelines, and more.
→
Explain how configurations work.
→
Automate repetitive tasks for your team.
Since it's powered by the MCP protocol, you choose which LLM to connect. Any model that supports tool-calling can be used, giving you full control.
AI in the experiences you build.
With Workflows, you can connect to external AI APIs and bring intelligence directly into the apps. This can happen in two ways:
1. Direct AI — real-time user features
For example:
→
Chat with the catalog to discover new content
→
Get live answers about a show, team, or event
→
Receive personalized suggestions or dynamic UI elements
These features are powered by Workflows running in the apps or in real time on the server — set up just like any other functionality.
2. Indirect AI — backend enrichment
For example:
→
Automatic subtitling and translations
→
Metadata enrichment, summaries, or categorization
→
Generating alternative assets like images or text snippets
These processes typically run server-side, often triggered by content ingestion workflows, so enriched assets are ready when audiences arrive.
Try our Color Assistant.
Assistant
Hello! Let me help you set the right vibe for this space. What kind of theme would you like?
""
Using Workflows we've enabled tools for the LLM to control background and text colors. Simple, but fun!
Flexible by design.
Whether it’s Doorman and the MCP server helping your team inside the EMS, or external AI services shaping what users see and feel, Doors makes AI practical and integrated. You own the workflows, the models, and the data — and you decide whether AI runs in real time, behind the scenes, or both.