Former Meta Neural Band Lead Seeks Funding for Flourish at $2.5 Billion Valuation

Thomas Reardon, one of the key people behind the Neural Band project at Meta, is raising money for a new startup, Flourish. The company wants to work on energy-efficient artificial intelligence and is discussing a round at a $2.5 billion valuation.

The round and the valuation

According to people familiar with the talks, Flourish is in discussions with investors about a new round of funding. The valuation under discussion is $2.5 billion, but the deal has not closed yet, so the terms may change. For a young company this is a very ambitious bar: the market is effectively betting on the team and the idea, not on a large-scale product shown publicly.

Little is known about the product so far, but the startup is already trying to take on one of the most expensive and painful problems in the AI industry. Flourish is described as a company that will build energy-efficient AI. It has not been disclosed publicly whether this means a new model architecture, optimization tools, a software stack for inference, or a pairing with specialized hardware. But the focus is already clear: this is not another interface on top of someone else's API, but an attempt to cut the cost of compute. At a time when models are getting stronger and costing more to serve, such a thesis looks attractive to investors on its own.

Who Reardon is

The founder matters here as much as the idea. Thomas Reardon led the work on Neural Band at Meta, a wristband neural interface that reads muscle signals and translates them into commands for a computer. Before that, he founded CTRL-Labs, a company at the intersection of neuroscience and computing, which Meta later acquired. After the deal, Reardon oversaw neuromotor interfaces and device interaction at Reality Labs.

Reardon also currently works as a venture partner at Lux Capital. That explains why Flourish has strong access to deep tech investors from the start, and why the project looks less like yet another AI app built on someone else's model and more like an attempt to enter a more fundamental layer of the market. When a startup is run by someone with experience in interfaces, neuroscience and computing systems, investors are ready to look at it differently and are more tolerant of early product uncertainty.

Why it matters

People familiar with the matter say Lux Capital and Google Ventures have already agreed to become major investors in the deal. Even without many public details, this is an important signal: at an early stage, a startup can get not just money but confirmation of interest from funds that usually look at complex technological bets. For the market, it looks like a test of the hypothesis that AI energy consumption has become an investment thesis in its own right, rather than just a nice phrase in pitch decks.

• an early-stage project is going for a late-stage unicorn-level valuation
• the founder fits a difficult technical field almost perfectly
• investors are betting not on a chatbot but on infrastructure or model efficiency
• AI energy consumption is becoming an investment thesis in its own right

Against the backdrop of the generative AI boom, this looks logical. The most obvious ideas are already taken by big players, and new money is increasingly flowing to companies that tackle narrow but expensive constraints of the market: compute, bandwidth, the cost of running models and the resilience of infrastructure. Not long ago investors were more willing to back applied AI services with a quick path to revenue, but attention is now shifting toward foundational technologies. If Flourish can show real technological novelty, a $2.5 billion valuation will stop looking like a pure bet on a name.

What it means

The Flourish story shows where investor interest is moving: from yet another AI service to technologies that make AI itself cheaper and more practical. If Reardon's project really finds a way to noticeably improve energy efficiency, it would benefit not only individual models but the entire AI chain, from infrastructure to end products, where the cost of compute increasingly and directly affects strategy and business margins.

*Meta has been designated an extremist organization and is banned in Russia.