Vancouver-based Variational AI has raised US$5.5 million in an oversubscribed funding round and plans to double its nine-employee staff count, CEO Handol Kim told BIV this afternoon.
"We haven't had the gas in the tank to fund business development and go to market," Kim said.
"It's now funding us, going and commercializing and getting more than just a few customers. It really supercharges our revenue and customer traction."
Nimbus Synergies led the round, with participation from Merck Global Health Innovation Fund, Quimby Investments, Threshold Impact, and Defined Capital. Existing investors Flying Fish, A&E Investment, and Nepenthe Capital participated.
Kim said money will go to increase his company's compute capacity, particularly by getting more graphics processing units (GPUs).
His company has dedicated servers at an off-site data centre and capacity from those servers will increase, he said.
While some artificial intelligence (AI) platforms create pictures, and others generate answers to questions or spit out long essays in seconds, Variational AI's Enki generative-AI platform uses the language of chemistry, Kim explained to BIV in 2023.
Variational AI's generative-AI platform can produce, for example, a recipe for a molecule that is of a certain weight and has a specific number of rotatable bonds or has the ability for certain pharmacokinetic interactions, he said.
The company last year announced a partnership with multinational drug developer Merck.
Drug developers and biotechnology companies have been increasingly using AI for early-stage drug discoveries.
Small-molecule drug discovery continues to rely on older, non-generative AI models focused on finding hits from within large libraries of virtual molecules, Variational AI said in a press release today.
While effective, this approach often limits novel discoveries and still requires extensive optimization, it added.
Generative models, in contrast, are extremely effective in domains such as language and images but have yet to make a similarly deep impact on chemistry, the company said.
“The Variational AI scientific team has spent years quietly building a foundation model constrained to small molecule drug discovery that uses orders of magnitude less compute and data than foundation models in other domains” said Jason Robertson, managing partner at Nimbus Synergies.
“This foundation model marks a paradigm shift in drug discovery enabling better, more reliable, and cost-effective drug development, which ultimately improves patient outcomes and reduces healthcare costs,” he added.
