Building Medicines from Scratch: How AI Proteins Plans to Redefine Drug Discovery
In a field where most drug developers start with what biology already provides, AI Proteins is taking a very different route. The Boston-based biotech startup designs medicines from scratch, using artificial intelligence, de novo protein design, and synthetic biology to create miniproteins that have never existed in nature. It is an engineered approach to therapeutics—one that promises better precision, finer control, and entirely new treatment possibilities.
“We start our design process with a target product profile and work backward to create the exact molecule needed,” CEO and Founder Dr. Chris Bahl told MoveTheNeedle.news. It sounds straightforward, but it represents a clear break from the past forty years of drug development. Instead of screening natural molecules to see what might work, AI Proteins aims to build exactly what a disease requires.
The company's bold vision now comes backed by significant capital. On 19 November 2025, AI Proteins announced a US$41.5 million Series A round, led by Mission BioCapital and Santé Ventures, with participation from Lightchain Capital and Cobro Ventures. It is one of the biggest financings to date in the growing field of AI-driven protein therapeutics.
From Nobel-Winning Science to a New Biotech Modality
After completing his PhD in biochemistry, Dr. Bahl joined the Baker Lab in 2012 as a postdoctoral researcher. This University of Washington lab is widely seen as one of the most influential laboratories in modern biology—David Baker earned a Nobel Prize in 2024 for groundbreaking contributions to protein folding and de novo design.
As part of this group, which was redefining what proteins could be, Dr. Bahl worked on early examples of miniproteins: compact, stable proteins engineered entirely from first principles. This was the work that convinced him the world was on the edge of a new therapeutic era.
In a 2024 interview with Endpoints News, Dr. Bahl reflected on that turning point: “The idea that you could build a therapeutic molecule that had never existed in nature… that changed everything.” The experience shaped both his scientific philosophy and his career trajectory.
After leaving the Baker Lab, Dr. Bahl became founding faculty at the Institute for Protein Innovation (IPI) and held appointments at Boston Children’s Hospital and Harvard Medical School. At IPI, he built a team of protein engineers who would eventually follow him into AI Proteins in 2021. He hadn’t planned to start a company—he thought he would stay in academia—but early investors convinced him that the best way to realize the potential of de novo protein therapeutics was to lead the company dedicated to it.
Why Miniproteins Matter in the Future of Drug Discovery
Dr. Bahl explained that miniproteins sit in an unusual place in the landscape of therapeutic modalities. They are larger and more structured than peptides, which often degrade quickly in the body, but far smaller and more engineerable than antibodies, which are shaped by evolutionary constraints.
Designed entirely from scratch, miniproteins—typically under 100 amino acids—are small enough to fold predictably yet large enough to carry out sophisticated functions. Because they are synthetic, they can be fine-tuned for stability, pharmacokinetics, target affinity, and manufacturability in ways that natural proteins cannot.
Dr. Bahl describes the process as starting with an architectural specification sheet. Rather than beginning with an existing molecule and trying to improve it, AI Proteins uses its platform to generate thousands of fully synthetic candidates designed to meet precise performance requirements. Many of these molecules possess structural features or behaviours that do not exist anywhere in biology—and some combine functions that evolution never placed in a single molecule.
This, he suggests, is the real promise of AI-enabled de novo design: the ability to engineer therapeutic behavior rather than merely observe it.
A Platform Built to Create, Not Just Discover
AI Proteins’ engine combines machine learning, synthetic biology, and high-throughput experimentation. The process begins with computational design, where AI models generate miniprotein structures. These designs move directly into robotic synthesis and testing, allowing the team to evaluate vast numbers of molecules in parallel.
The result is a platform built for speed. Instead of investing years into modifying a single antibody or peptide, the company can generate, test, refine, and iterate on entirely new proteins in a fraction of the time.
To preserve the value of this engine, AI Proteins has adopted a hub-and-spoke structure, highlighted during the Series A announcement. At the center is the core platform, which the company regards as its long-term asset. Around it sit a constellation of programs and subsidiaries—each focused on its own therapeutic area but all benefiting from the same design technology.
“It’s a way of scaling innovation without diluting the platform’s value or tying the company’s future to a single program,” Dr. Bahl said. “The platform is our raison d’être—this ability to generate molecule after molecule.”
Partnerships That Signal Momentum
AI Proteins’ rapid progress has not gone unnoticed. In December 2024, the company entered a collaboration and option agreement with Bristol Myers Squibb (BMS) worth up to US$400 million in milestones. Under the agreement, AI Proteins designs miniproteins against targets selected by BMS. If the resulting molecules meet predefined criteria, BMS can license them—and up to two additional programs—for clinical development and commercialization.
This early validation from a major pharmaceutical company underscores the growing belief that de novo protein therapeutics may represent the next major drug modality.
The company’s internal achievements are equally notable. In a 2025 disclosure, AI Proteins confirmed it had already generated molecules against more than 150 targets, with several demonstrating in-vivo proof-of-concept. These milestones are unusually rapid for a biotech founded just four years earlier and reflect the velocity of a platform-driven design model.
A Pipeline Taking Shape
Although AI Proteins has not publicly released a full pipeline chart, the company has confirmed active programs in oncology and inflammation. Its oncology work ranges from hematologic cancers to solid tumours, areas where miniproteins offer advantages: the ability to hit multiple antigens, penetrate tumours more effectively, and deliver payloads with far greater precision than conventional biologics.
The company is also advancing programs in inflammatory and metabolic diseases, where the specificity and tunability of miniproteins could unlock new therapeutic options. While AI Proteins has not disclosed a timeline for entering the clinic, the Series A financing and BMS partnership indicate that multiple programs are moving toward IND-enabling studies.
A Pragmatic First Step—Then Much More
Despite the futuristic implications of designing drugs from first principles, Dr. Bahl insists that AI Proteins’ early candidates will be scientifically conservative and low-risk. The first generation, he says, “won’t be very sci-fi.” They will behave like familiar biologics, making them easier to advance through regulatory pathways and easier for partners to adopt.
Once those molecules establish safety and efficacy, the company intends to push further into synthetic territory—creating miniproteins with capabilities far beyond what evolution has produced. Dr. Bahl also envisions long-term applications outside therapeutics, including agriculture, food technology, and industrial biology.
His ambition is expansive: within ten years, he hopes AI Proteins will have contributed to 50 new medicines.
The Bigger Picture for AI-Driven Drug Design
What makes AI Proteins compelling is not just its platform or partnerships but the pedigree and perspective behind it. Dr. Bahl emerged from the Nobel Prize–winning Baker Lab, a scientific environment that proved proteins could be designed as predictably as engineered systems. AI Proteins is now trying to translate that philosophy into a new category of medicines—ones defined not by the accidents of evolution, but by deliberate human design.
If the company succeeds, miniproteins could become one of the most important therapeutic tools of the coming decades, reshaping how the biotech industry approaches drug discovery.