Otter Debuts AI Meeting Assistant OtterPilot to Summarize Discussions
Meeting transcription and collaboration platform Otter.ai has introduced an AI assistant named OtterPilot this week. OtterPilot employs AI to automate meetings, including taking notes and generating a summary of the discussion, including slides and other images, that is then shared with those invited to the meeting or other designated recipients.
“We have AI in every part of our lives. For instance, I’ve been super impressed with Tesla’s Autopilot. It takes me from home to the office almost completely automatically. Otter CEO Sam Liang said in an interview with Voicebot. “We’re calling it ‘OtterPilot’ because We want to use Otter to drive your meeting automatically the same way.”
The automation includes automatically joining meetings scheduled on a Microsoft or Google calendar and real-time transcription and note-taking where people can add comments and action items. OtterPilot will also screenshot any slides or other presentation images, inserting them into the notes. OtterPilot’s centerpiece is its automated meeting summary, which condenses the main topics and decisions into a highlight reel with hyperlinks to the part of the transcript mentioned. The AI model for the automated summary is different from the one used in the meeting outline feature introduced last year.
“The summary is more informative and more accurate, and it can extract information from slides. Even the notetaking is much more advanced than it was a few months ago or a year ago,” Liang said. “It responds to feedback, like when people correct the transcript or use new terminology. The system learns and improves. In the conversations, you can actually see the accuracy of the transcript and summary improve.”
Though Liang didn’t refer to OtterPilot as generative AI, it does share some traits on the surface. The use of natural language processing and deep learning models to summarize a meeting and its adaptive nature feel a lot like what OpenAI and its rivals are showcasing. Generative AI is already rapidly spreading among other meeting AI service providers like Supernormal, which recently raised $10 million.
“Everyone is talking about ChatGPT, but we’re focused on improving enterprise productivity,” Liang said. “We are investing a lot more to build a ChatGPT-type service for each enterprise, so [the AI] really understands their projects and their team. Our long-term planning is to build personal models for each person with their own style and their own terms.”
OtterPilot arrives as the company reaches a milestone of one billion meetings transcribed since its 2016 foundation. Demand for AI meeting assistants, in general, exploded as the COVID-19 pandemic pushed more offices to operate work-from-home environments. Startups are raking in cash, with examples ranging from Airgram’s $10 million and Xembly’s $15 million to the $70 million investment in Kore.ai or the $62 million Speechmatics raised last summer.
“It’s a really exciting time, with lots of new use cases. We are really well positioned because of the number of meetings we have transcribed. Most startups don’t have this data,” Liang said. “We’re pretty proud to build at the same level as Google and Microsoft at a fraction of the size.”