Designing the vaccine before the virus mutates - and now aiming the same AI at diseases with no cure.
Most vaccines are built to match the pathogen we can see today. Baseimmune's wager is that the more useful question is what the pathogen will look like tomorrow. The London biotech runs a proprietary deep-learning platform that designs synthetic antigens - the active ingredient that teaches the immune system what to attack - by assembling the parts of a pathogen most likely to trigger broad, durable protection.
The company was founded in 2019 by three researchers - Joshua Blight, Ariane Gomes and Phillip Kemlo - who met through work at the University of Oxford's Jenner Institute, one of the world's best-known vaccine research centres. Two came from the biology bench; one built the software. That mix is the point: Baseimmune treats antigen discovery as a computational problem first and a wet-lab problem second.
The practical claim is specific. Rather than reacting to each new variant after it spreads, the platform models how a pathogen is likely to mutate and designs an antigen around the conserved, functionally essential regions it cannot easily change. The company has said its model anticipated major COVID-19 variants, including Alpha and Delta, from limited early data - a proof point it uses to argue the approach generalises beyond any single disease.
"When we started, we had an idea. Now, we have data that shows we can actually make better vaccines with broader protection and tackle complex pathogens in a way that wasn't possible before."
Ariane Gomes - Co-Founder & Chief Scientific OfficerIn vaccine development, the expensive failures happen early. The hardest decision is not how to manufacture a vaccine but which target to build it around. Pick the wrong antigen and years of clinical work can be spent chasing protection that never arrives - a particular problem for pathogens that mutate quickly or hide from the immune system.
Baseimmune's answer is to move that decision to where a computer can help. Its platform integrates deep-learning antigen design, structural modelling and experimental screening, letting the team simulate and rank candidate antigens before committing to the bench. The goal is not to skip the lab, but to arrive there with fewer dead ends.
That is also how it differs from conventional discovery inside large pharma, which still leans heavily on empirical trial and error. And unlike some AI-biology peers focused narrowly on protein structure prediction, Baseimmune is oriented around a downstream product: an antigen designed to provoke a specific, protective immune response.
A proprietary deep-learning system that designs synthetic antigens by combining the pathogen components most likely to trigger broad protection, backed by structural modelling and lab screening.
Since 2019A preclinical, variant-resilient COVID-19 vaccine candidate designed to stay effective as the virus changes.
PreclinicalA preclinical program aimed at the malaria parasite - one of the field's longest-standing hard targets.
PreclinicalA preclinical veterinary vaccine candidate against a virus with major consequences for global food supply.
PreclinicalA multi-pathway immunotherapy applying the antigen platform to idiopathic pulmonary fibrosis, with proof-of-concept readouts expected in 2026-2027.
Announced 2026Baseimmune follows a familiar biotech shape with an AI twist. It uses its platform to build a pipeline of proprietary candidates, works to de-risk them through preclinical development, and expects to earn revenue by out-licensing assets and co-developing with larger pharmaceutical and animal-health companies - rather than manufacturing and selling vaccines itself.
That places it in a fast-growing corner of the market where computational biology meets drug discovery, alongside players such as Evaxion Biotech, Generate Biomedicines and Nabla Bio, and in competition with the in-house discovery engines of established vaccine makers. In fibrosis, the incumbents are approved drugs like pirfenidone and nintedanib, which slow lung-function decline but do not reverse it.
The 2026 expansion into fibrosis is the clearest signal of the company's ambition: if antigen design is genuinely a general tool for instructing the immune system, its value is not confined to infection. The tell to watch is data - the proof-of-concept efficacy readouts due across 2026 and 2027.
| Round | Amount | Year |
|---|---|---|
| Seed | $4.8M | 2021 |
| Series A | $11.3M | 2024 |
Oxford Jenner Institute researcher turned company builder, leading Baseimmune's strategy and pipeline direction.
Chief Scientific Officer and the scientific voice of the platform, focused on translating computational design into real immune protection.
The software engineer who built the antigen-design platform - the engine that turns pathogen data into candidate antigens.
The three met while pursuing doctoral research at Oxford, where the vaccine method they worked on reached clinical trials faster than any prior candidate at the Jenner Institute. That combination - vaccine biology plus software - is the expertise Baseimmune was assembled around.
Blight, Gomes and Kemlo spin the company out of research connections at Oxford's Jenner Institute.
Raises a $4.8M seed round to build the antigen-design platform, with early support including a former Moderna executive.
The platform demonstrates it can anticipate major COVID-19 variants from limited early data.
Recognised among promising startups while advancing malaria and African swine fever programs.
Closes an oversubscribed round led by MSD Global Health Innovation Fund and IQ Capital.
Launches a lead immunotherapy program targeting idiopathic pulmonary fibrosis, extending the platform beyond infectious disease.
It designs synthetic antigens - the active ingredients in vaccines and immunotherapies - using a proprietary deep-learning platform, originally for fast-mutating pathogens and, since 2026, for chronic diseases like fibrosis.
It was founded in 2019 by Joshua Blight, Ariane Gomes and Phillip Kemlo, who connected through research at Oxford's Jenner Institute.
A $4.8M seed round in 2021 and an $11.3M (about GBP 9M) Series A in February 2024 led by MSD Global Health Innovation Fund and IQ Capital.
Rather than designing a vaccine for the current form of a pathogen, its AI predicts likely mutations and builds antigens from conserved, essential components, aiming for broader and more durable protection.
Alongside preclinical vaccine candidates for coronavirus, malaria and African swine fever, it launched a lead fibrosis (IPF) immunotherapy program in 2026, with proof-of-concept readouts expected in 2026-2027.