Otter.ai launches conversational knowledge engine

Otter.ai said April 28 it launched a Conversational Knowledge Engine, a new product layer that turns meeting transcripts into searchable company knowledge. (otter.ai) The company said the system connects conversations across teams and time, then uses that record to answer questions and trigger follow-up work. Otter tied the launch to three product changes: AI Chat Connectors, an expanded Model Context Protocol server, and Otter for Desktop. (otter.ai) The new connectors let users pull live data from Gmail, Google Drive, Notion, Jira and Salesforce into Otter AI Chat. Otter said Microsoft Outlook, Teams, SharePoint and Slack connectors are coming next. (otter.ai) (techcrunch.com) Otter said users can ask whether a follow-up email was sent, surface a budget confirmation from Gmail, push meeting summaries to Notion, or draft and send a Gmail message from inside Otter. (otter.ai) (techcrunch.com) Model Context Protocol, or MCP, is a standard for connecting artificial intelligence tools to outside data sources and services. Otter said its AI Chat Connectors make the company an MCP client, while its earlier enterprise push let organizations build custom MCP connections to Otter data. (techcrunch.com) (otter.ai) That shift moves Otter beyond transcription and summaries, a category that is getting crowded with products from Read AI, Fireflies.ai and Fathom. TechCrunch reported those rivals are also trying to become broader workspaces that pull in outside business data. (techcrunch.com) Otter also redesigned its assistant so it stays visible across the interface and answers questions based on the screen a user is viewing. The company is pairing that with desktop capture on Windows after bringing botless capture to its Mac app late in 2025, according to TechCrunch. (techcrunch.com) Chief executive Sam Liang told TechCrunch that many enterprise customers still prefer a note-taking bot that joins Zoom calls because it is visible to attendees and shares notes with the whole meeting. That leaves Otter selling both a broader knowledge layer and the older meeting-bot model at the same time. (techcrunch.com) Otter framed the new product as a system of record for what people say in meetings, then linked that claim to the tools where those decisions get carried out. The company is now betting that the transcript alone is no longer the product. (otter.ai)