Supernormal
Supernormal, a platform for AI meeting notes and voice agents, partnered with AssemblyAI to improve transcription accuracy and multilingual support—and take control of the market.
increase in free to paid conversion rate
tasks processed monthly
Identifying the virtual meeting opportunity
In 2020, Supernormal founder and CEO Colin Treseler saw a huge market opportunity with the rise in the popularity of virtual meetings.
But what happens when key decision-makers have to miss meetings? Of course, notes or emails can be circulated, but that person misses the nuances and depth of the discussion. Even if everyone can attend the meeting, Treseler continues, it can be distracting and difficult for attendees to take meaningful notes while actively participating in the conversation.
Treseler and his team started to look at how AI could solve this problem—and Supernormal was born.
Building the foundation for AI meeting intelligence
The team's first goal was to automate virtual meeting transcription so that all conversations would be transferred into an easily digestible, shareable format. Then, the team needed to take this transcription and generate additional insights from the meeting.
However, their legacy speech-to-text provider couldn't transcribe meetings accurately enough to make these meeting notes usable, either as notes themselves or to spur further workflows, like creating action items or meeting summaries.
That's when Treseler discovered AssemblyAI:
Launching AI-powered features
Increased transcription accuracy also allowed the team to launch additional AI-powered features that rely on this transcription data, like its popular Ask Norma feature. Ask Norma is a meeting assistant that Supernormal users can talk to and bring insights to life with prompts like:
- Norma, remind me what we said about that earlier.
- Norma, what's another way to think about this?
- Norma, can you recap the latest team meeting?
- Norma, what ideas do you have?
Today, Supernormal has seen its customer base—which includes names such as QuantaData, Tribe AI, Fiat Growth, and Shima Captial—explode. The platform now processes between 10 and 15 million tasks based on its users' virtual meetings every month.
Doubling conversion rates
The team also saw significant results in their metrics:
Expanding internationally with multilingual support
And as Supernormal's business expanded internationally, so did its need to include highly accurate multilingual transcription in its platform. A large Brazilian airline, for example, reached out to Supernormal to test their product in large-scale virtual meetings. The company was so impressed that Supernormal "had an immediate demand and huge influx of requests to customer support around being able to support more languages," Treseler explains.
Launching voice agents and future growth
Most recently, Supernormal launched Voice Agents, an addition to its AI meeting platform that allows its customers to create customizable conversational agents that can handle routine conversations like inbound sales, customer support, employee surveys, and more.
This feature builds on Supernormal's key mission—boosting productivity so customers can focus on more meaningful work.
Start building with $50 in free credits
Start building with Universal-Streaming and create voice agents that feel natural, responsive, reliable, and genuinely helpful.
A partnership built on support and scale
Siro
Siro, an AI-powered coaching platform for field sales, integrated AssemblyAI to accurately transcribe and identify speakers in the field, leading to significant downstream benefits for its customers.
Delphi
Expertise can be transformative, and Delphi is pioneering a revolutionary approach to knowledge sharing. By harnessing advanced AI technology, the company is creating digital clones of thought leaders, entrepreneurs, and experts, making their insights accessible to a global audience 24/7.
Turn voice data into unparalleled product experiences
Partner with the leader in Speech AI to build powerful products with breakthrough industry impact.
