The promise of AI and large language models (LLMs) is the ability to understand increasingly wider amounts of context and make sense of that information easily, so it makes sense that we’re seeing a bunch of companies trying to make wearable hardware so that people can use AI in their day-to-day lives.
The latest entrant in this space is Bee AI, and it has raised $7 million in a round led by Exor to build out its wearable AI assistant that listens to you to learn more about you, take notes, surface contextual reminders and build lists. The company also has a companion Apple Watch app.
The amount being disclosed today includes pre-seed funding of $1.5 million that the startup had raised previously. Greycroft, New Wave VC, Banana Capital, and Brian Bedol (an investor and TV exec who previously founded a couple of sports networks) also participated in the new round.
Co-founder and CEO Maria de Lourdes Zollo told TechCrunch that while the core focus of Bee AI is the software powering the assistant, the company built a wearable so the app doesn’t need to constantly take control of a user’s phone mic.
The device and app can be used to do various tasks as mentioned earlier, but the startup is ambitious. De Lourdes Zollo said the company wants to give each consumer a “cloud phone” — essentially a mirror of your phone with access to your accounts and notifications. At the moment, some of the features in early testing include the ability to read your notifications and get reminders about important messages and events, write emails or tweets, and get shopping suggestions on demand.
The device currently just has a mute button to stop recording, but the company is exploring ways to also use the button to trigger commands.
The opportunity and roadmap
Given that generative AI is so new, there are still questions about its ability to output reliable information, so there’s some skepticism in the space Bee AI is entering. Startups like Rabbit have tried using AI agents that can traverse an interface on your behalf to do various tasks. However, as early reviews and demos suggest, the process doesn’t work reliably yet.
Still, there are a few startups working on the problem in hopes that they will be the first to get it right: A16z-backed Limitless and Friend are both building wearables that promise to do similar things as Bee AI, though the use-cases they are tackling are slightly different. For her part, de Lourdes Zollo thinks that AI agents will improve as new models are released, and said Bee AI is taking conservative approach by focusing on doing just a few tasks.
De Lourdes Zollo founded Bee AI with Ethan Sutin (CTO), who she worked with at video chat app Squad, which he founded with Esther Crawford (also an angel investor in Bee AI). They also worked at Twitter, where Sutin was an engineering lead, and de Lourdes Zollo helped the platform ship Twitter Spaces.
The company’s investors seem to be confident of the team’s pedigree. Ian Sigalow, managing partner at Greycroft, said that he saw great potential in the team and decided to invest because of that. “I generally invest in great founding teams. With Bee AI, you have unique team members who are savants of engineering. Many of them have worked at a company like Twitter and shipped products to millions of users. I think that’s a great strength.” he told TechCrunch.
Sigalow also thinks there could be great opportunities in building a product that can do solid handoffs between hardware and the cloud if you train large language models well.
There are bound to be some privacy concerns about a device that listens to you all the time to function. The product is in beta at the moment, and currently also uses what people in the user’s vicinity say to provide more context and improve the model’s learnings about the user. However, before launch, the company aims to stop using all non-users’ voices if they have not given verbal consent to be recorded.
And, Bee AI claims its platform doesn’t store any audio recordings and instead only uses transcripts to learn more about the user.
Bee AI will be priced at $49, along with a $19 per month subscription. The company aims to start taking orders before Black Friday.