Apoha emerges from stealth with $36M Series A to build AI for new materials
What's the deal? ApohaDealroom has a profile for this one. Try Dealroom →, a London- and San Francisco-based deep tech startup, has emerged from stealth with a $36M Series A to build AI models that can design new substances — from proteins to paints to pharmaceuticals. The round was led by European VC firm Singular, with participation from Draper Associates and existing seed investors Redalpine, Seedcamp, WilbeDealroom has a profile for this one. Try Dealroom →, and Nucleus. Apoha also has a grant from Innovate UK. It did not disclose its valuation.
The company's core bet: most AI models learn from text or images, but materials science needs data about how substances actually behave. Apoha has built lab hardware that suspends a pin-head-sized sample in liquid, applies tiny physical stresses, and records the wave patterns that ripple back. Those patterns yield over 1,000 numerical descriptors in minutes — versus the days or weeks conventional lab tests require.
Apoha calls the approach "liquid intelligence." It converts raw wave recordings into what it calls "behavioural embeddings" — numerical fingerprints AI models can learn from. The company's first commercial product, VIBE (Variations in Inter-facial Behaviour Under Excitation), can predict whether a drug will hold together in the body, whether a plant-based protein will tear like chicken, or how a coating will wear over time.
Why now? Cofounded in 2021 by Shamit ShrivastavaDealroom has a profile for this one. Try Dealroom →, a mechanical engineer with post-doctoral research at Oxford, and Anshika SrivastavaDealroom has a profile for this one. Try Dealroom →, a former Goldman SachsDealroom has a profile for this one. Try Dealroom → banker, Apoha spent years building proprietary hardware and generating a dataset that didn't previously exist at scale. That groundwork is now mature enough to serve paying customers.
"Machines have learned to see what matter looks like and to read what we say about it," Srivastava said. "They have not learned to taste, smell, or feel matter — to perceive how a drug dissolves, how a flavour holds, how a material wears. That is the layer we are building."
What could go wrong? Apoha is creating an entirely new data category, which means it must convince industries to trust measurements that have no established benchmark. Scaling proprietary lab hardware is capital-intensive, and the company will need to prove its predictions hold up across diverse real-world applications — from pharma to food — before it can command broad adoption.
The signal: Singular leading a $36M Series A for an early-growth startup whose core asset is proprietary lab hardware — not software alone — underscores a broader investor shift toward deep tech bets that generate entirely new datasets rather than repackaging existing ones. The backing from Seedcamp and Redalpine since the seed stage, now joined by Draper Associates from the US, suggests growing transatlantic conviction that AI's next value frontier lies in physical-world data modalities that incumbents cannot easily replicate.
Read more: Fortune
