Apoha Unveils "Liquid State Intelligence" with $36 Million Funding

Apoha, a deeptech company pioneering what it calls "Liquid State Intelligence," today announced its emergence from stealth with a significant $36 million funding round. Singular led the investment, with participation from Tim Draper’s Draper Associates and continued backing from seed investors Redalpine, Seedcamp, Wilbe, and Nucleus. This is complemented by grant funding from Innovate UK, the UK’s national innovation agency.

For decades, the scientific community has faced a fundamental gap in understanding molecular behavior. While it was possible to determine a molecule's sequence and structure, a clear vision of how it actually behaved under real-world conditions was missing. The only available information consisted of limited and specific measurements, conducted under controlled lab conditions. This uncertainty has led companies to make billion-dollar decisions with a high degree of risk, such as introducing drugs into clinical trials without a full understanding of their efficacy in patients, launching food products without a clear perception from the end consumer, or using new materials without a deep knowledge of their everyday performance.

"Liquid State Intelligence": The Missing Piece for Physical-World AI

As the frontier of artificial intelligence extends beyond language and code to act on the physical world, a further challenge emerges. Machines have learned to "see" and "read," but they have not yet learned to "feel" matter: to perceive how a drug responds to real-world conditions, how a flavor is perceived, or how a material functions in everyday contexts. Vision and language have been digitized at scale, but the behavior of matter has remained an unknown.

This is where Apoha's "Liquid State Intelligence" comes in. The company creates the missing data layer for physical-world AI: large-scale empirical data about how matter behaves. This approach enables more powerful reasoning and prediction, facilitating autonomous scientific discovery. Apoha defines this new category of molecular science, alongside sequence and structure, as "Liquid State Intelligence." By building this foundational data class of molecular behavior, Apoha aims to redefine the processes of discovering and designing medicines, food, and materials. The science behind Apoha traces back to 2008, when founder and CEO Shamit Shrivastava began researching a problem left unresolved by the Nobel Prize-winning Hodgkin-Huxley model of nerve signal transmission: the physics of what happens at the boundary where matter meets liquids. His discovery, that these boundaries carry a measurable record of molecular interactions, has since been cited over 1,500 times.

From Research to Product: The VIBE® Platform and its Applications

In 2021, Shrivastava co-founded Apoha with Anshika Srivastava, COO and former Executive Director at Goldman Sachs, to transform his research into a global platform for measuring molecular behavior at scale. Today, the company holds over 60 patents covering hardware, software, data, and AI models. Apoha's first product is VIBE® (Variations in Inter-facial Behaviour Under Excitation), an empirical readout of how a molecule, material, or formulation behaves under real-world conditions.

To generate this data, Apoha's platform takes a tiny sample, suspends it in liquid, applies a controlled series of stresses, and captures the wave patterns the molecule generates in response. Recorded in real time, these wave patterns translate into over 1,000 empirically measured descriptors of behavior. A single VIBE® readout can simultaneously resolve what conventional approaches measure one property at a time. Within minutes, a VIBE® readout can indicate whether an experimental drug will fail before it ever reaches a clinical trial, saving years of work and hundreds of millions of dollars per failed candidate. The same readout works across sectors such as food, materials, and any other domain where the behavior of matter determines a product's success.

Future Prospects and the Strategic Value of Empirical Data

The platform is already in commercial use. Joint research with Boehringer Ingelheim, a multi-year commercial partner, has shown Apoha identifying high-risk antibody candidates with greater than 90% precision from as little as 8 micrograms of material. For Ethris, the German biotech, Apoha is working on improving in-vitro to in-vivo correlation, predicting how lipid nanoparticles carrying mRNA will behave in animals. THIS, the plant-based food company, used Apoha’s technology to find a protein replacement for a product destined for supermarket shelves in record time. The company also partners with Somru BioSciences and multiple Fortune 500 companies across the pharmaceutical, food and beverage, and materials sectors.

Shamit Shrivastava, CEO and co-founder at Apoha, stated: "Liquid State Intelligence took 15 years of science and 5 years of company-building to bring to life. There is no shortcut to this data class – it cannot be scraped from the internet, synthesized, or retrofitted from existing assays. It has to be measured. Where sequence gave us the language of biology and structure the language of design, Liquid State Intelligence gives us the language of behavior – what matter, molecules, and materials actually do – and we are the company building it." Raffi Kamber, Co-Founder and GP at Singular, added: "Apoha represents a new generation of European scientific companies where AI is not a future promise, but a practical tool already transforming how biology is done. For the first time in 25 years, we are back to creating genuinely new science, being commercialized by founders with drive and global ambition." The funds raised will be used to build Liquid State Intelligence into the foundational data class for molecular behavior, with applications in biologics, food, materials, and the next generation of physical-world AI. For companies evaluating AI solutions for complex workloads, the availability of high-quality empirical data like that generated by Apoha is a critical factor, directly influencing the effectiveness and reliability of models, whether they are deployed in cloud or self-hosted environments.