Colorado biotech Ambrosia Biosciences closed a $100 million Series B round on March 31, 2026, according to the Fierce Biotech fundraising tracker. The round drew a roster of crossover and healthcare-focused investors — Blue Owl Healthcare Opportunities, Redmile Group, Deep Track Capital, BVF Partners, Boulder Ventures, Janus Henderson Investors, Samsara BioCapital, and one unnamed institutional backer. It is a substantial vote of confidence for a team that has never published a single molecule in human testing. That is exactly the problem.
Ambrosia was founded in 2024 by former Array BioPharma scientists whom Pfizer laid off when it closed its Boulder, Colorado research facility in June of that year. Array, which Pfizer acquired for $11.8 billion in 2019, was, in the words of Kyle Lefkoff, Ambrosia's executive chairman and Array's founder, "among the best small molecule drug discovery teams on the planet." The bet is that the same team can solve a problem that has beaten well-funded competitors for a decade: an oral pill that works as well as the GLP-1 injections dominating obesity treatment today.
The pitch is coherent. The data to support it does not yet exist.
The credibility gap is specific. In August 2024, Nick Traggis, Ambrosia's founder and CEO, told Chemical & Engineering News that the company would present its first preclinical data sets near the end of 2025 and file its first investigational new drug application with the FDA around the end of 2026. As of March 31, 2026, that preclinical data has not appeared — no press release, no conference abstract, no update on the company website. The $100 million Series B closed on that same date without any disclosed preclinical proof of concept. Traggis did not respond to a request for this article, but in an earlier exchange he said the data presentation was always conditional on internal milestones being met and that pharma companies routinely keep IND-enabling results private until they file. That is a defensible explanation. It is also a $100 million gap in the public record for a company whose entire story rests on being further along than it appears.
The competitive window is not waiting. Eli Lilly's oral GLP-1 candidate orforglipron could receive FDA approval within weeks — the agency's decision is due on or before April 10, 2026, according to BioSpace's regulatory calendar. Lilly's Phase 3 ATTAIN-2 trial showed participants lost an average of 22.9 pounds, or 10.5 percent of body weight, on the highest dose. Structure Therapeutics, another oral GLP-1 developer, reported positive Phase 2b topline data in December 2025 with zero cases of drug-induced liver injury or persistent liver enzyme elevations. Structure's drug, aleniglipron, was designed as a biased GPCR agonist — the same receptor family Ambrosia is targeting. Ambrosia is still in IND-enabling studies. It is at least a year, and likely more, from filing. By the time it reaches a Phase 1 trial, Lilly's oral GLP-1 may already be on the market.
There is also an inconsistency in the public record that Ambrosia has not resolved. The Fierce Biotech fundraising tracker lists Phase 1 clinical work in Australia beginning in early 2024. But the company's own statements to Chemical & Engineering News in August 2024 described the IND filing as a goal around the end of 2026 — a timeline inconsistent with an Australian Phase 1 starting in 2024. Either the company is further along than its most recent public statements indicate, or the tracker entry reflects stale data that was never corrected. The company website makes no mention of clinical-stage programs. Clarification from Ambrosia would resolve this; it has not provided it.
The scientific approach is credible. GLP-1 drugs currently on the market — Wegovy, Ozempic, Mounjaro — work by activating the GLP-1 receptor at the same site where the natural hormone binds. That orthosteric site drives both the beneficial insulin-regulation pathway and the beta-arrestin signaling linked to side effects including nausea, vomiting, and diarrhea. Ambrosia is targeting allosteric sites on the receptor instead — different positions on the protein surface where a molecule could bind and preferentially activate the desired pathway while avoiding the side-effect pathway. Septerna CEO Jeff Finer told Science Magazine that small molecules targeting novel binding sites might potentially offer improved tolerability. Boulder Ventures, covering the science in a post on small-molecule GLP-1 research, notes that Septerna published structural work identifying a novel allosteric pocket on the GLP-1 receptor, with 80 to 90 percent of the pocket's genetic sequence shared among GLP-1, GIP, and glucagon receptors — potentially enabling multi-receptor targeting. But actually finding a small molecule that works is another matter. As Traggis told Science Magazine in April 2025, "you are essentially starting from scratch to find an effective compound." The GLP-1 receptor has a large interface with the natural hormone that a much smaller compound needs to navigate.
Ambrosia's specific strategy is to develop separate small-molecule drugs for GLP-1, GIP, and amylin receptors, to be combined as multiple pills where dosing can be tailored per individual to minimize side effects. It is a flexible approach in theory. In practice, three pills a day is a different compliance bar than one injection a week.
The company integrates cryo-EM structural biology, biophysical binding assays, and AI-powered molecular design to accelerate GPCR drug discovery. The team of roughly 15 scientists is composed almost entirely of former Array employees who contributed to 30 IND filings, 25 new molecular entities, and 7 FDA approvals across their careers. That pedigree is the clearest evidence the company has to offer, and it is a legitimate one.
Pfizer's discontinuation of its oral GLP-1 candidate danuglipron in April 2025 — after a participant in a dose-optimization study experienced potential drug-induced liver injury that resolved after stopping the drug — created an opening that Ambrosia and others are trying to fill. Traggis has said the liver injury appears specific to danuglipron, not intrinsic to oral GLP-1 as a class. Structure's zero-event safety record in Phase 2a suggests he may be right. But that is Structure's data, not Ambrosia's.
Ambrosia has pedigree, a mechanistically coherent approach, and $100 million more than it had a month ago. What it does not have is a single public experiment showing its compound works in an animal, a cell, or a patient. The team that could solve oral GLP-1 may well be sitting in that Boulder facility. Whether their molecule will is the question the data was supposed to answer — and has not yet.
The FDA decision on Lilly's orforglipron is due within days. Ambrosia's answer, whatever it is, is still somewhere in a lab.
