Instruct your favorite AI agent to upload a PDF, provision a knowledge base, and autonomously launch a live, voice-led interactive presentation room for you using Seminara. The new Agentic Suite exposes full control to agents and developers via WebMCP, CLI, and multi-language SDKs.
When we built Seminara, the goal was simple: give anyone an autonomous, voice-led host that can deliver presentations 24/7, flip slides, and answer unscripted questions from attendees.
But clicking through a dashboard to set up every presentation defeats the purpose when you're already using Claude, Cursor, or autonomous agents for your workflows. Your agent should be able to spin up and manage presentation rooms for you.
That is why we built and open-sourced the Seminara Agentic Suite.
Here is what is included:
Model Context Protocol (MCP) & WebMCP: If you use Claude Desktop or navigate to Seminara in a WebMCP-enabled browser, your agent automatically discovers our presentation tools. You can hand it a PDF or pitch deck and tell it to create a live room with custom CTAs.
Multi-Language SDKs: Native client libraries for Node.js (seminara-sdk), Python (seminara), and Go (seminara-agentic-suite/go).
Terminal CLI: A lightweight CLI (seminara-cli) to spin up rooms, inspect quotas, and query machine-readable pricing straight from your shell.
Live WebRTC Delivery Engine: Every room runs our sub-300ms voice presenter, live slide sync, and in-session lead capture with instant webhook dispatch to HubSpot and Salesforce.
Everything is open-source under the MIT license on GitHub:
https://github.com/shivamselam/s...
You can try the live interactive presenter directly in your browser without creating an account:
https://seminara.online/live/demo
I’m around all day to answer technical questions on our WebRTC pipeline, knowledge grounding, or MCP implementation.