Biotech tools · AI protein design
Medium Biosciences
A small San Francisco company that says it designs custom affinity reagents with AI and delivers them in weeks. Founded in 2021, Y Combinator Summer 2021, and so far visible mostly through its own site, a protein engineering competition and a handful of data aggregators.
A tournament and a second place
In 2023 a group of protein engineers tried something unusual. They posted a set of enzyme datasets, invited anyone to predict how protein sequences would behave, and then asked the best teams to design new sequences that a lab would actually build and measure.
The organizers called it the Protein Engineering Tournament. An independent summary of the results says that “All teams engineered variants with improved activities, but TUM Rostlab came in first by generating the top score of an individual variant and the highest median of all qualifying variants according to the combined criteria for expression and thermostability.” The next line reads: “MediumBio and Marks Lab finished second and third, respectively.”
That line is one of the few outside descriptions of what Medium Biosciences does. It is also slightly odd. The same summary lists Medium Biosciences among the organizing institutions. The company was both a host of the competition and an entrant that placed second.
The problem
An affinity reagent is a molecule that sticks to one specific target. Antibodies are the classic example. Researchers use them to find a protein in a tissue sample, and diagnostic developers build tests around them. When a good reagent does not exist for a target, someone has to make one, and that has traditionally meant months of wet lab work.
Medium’s pitch is speed. Its site says it delivers “high-performance affinity reagents in weeks, not months.” The company is not the only group applying machine learning to protein design, and nothing found here compares its reagents with antibodies or with rivals. The claim is a promise, not a measured result.
Founders
Y Combinator’s listing names two founders, Mohamed Hassan Kane and Adil Yusuf. Databases disagree on titles. LifeScienceHistory lists Kane as founder and CEO. Getlatka lists Yusuf as founder and CEO. Crunchbase, per a search result title, calls Yusuf co-founder and CTO. No source read for this profile settles it.
Kane’s public profile shows a machine learning background. He was involved in MIT student groups, including a machine intelligence community he co-founded, and he lists an earlier student startup, Airspress, that he says failed to scale. His publications include work on protein function prediction and a text evaluation method called NUBIA. He also lists a patent, WO2024123733A1, made with Twist Bioscience, which he describes as using machine learning to improve several properties of a DNA synthesis enzyme. The patent page captured for this profile did not show inventor names, so that link rests on his own description.
No biography for Yusuf was found beyond his name on the tournament paper, the YC listing and the databases.
From enzymes to reagents
The company’s stated product has changed. Aggregator pages that carry older company text describe “a software platform that makes it really easy for industrial biotechnology companies to interface with machine learning models for enzyme engineering.” Today the site sells custom affinity reagents to research teams and diagnostic developers.
Timeline as far as the record goes:
September 2021: Y Combinator invests $125K, according to PitchBook and Tracxn. Late 2021 and late 2022: PitchBook lists two more early stage rounds with no amounts. 2023: the tournament pilot preprint appears with both founders among the authors. November 2025: the tournament results paper is published in the journal Proteins. 2026: the company site reports 25+ completed projects, 20+ unique proteins engineered and 10+ customers.
What is proven, and what is still claimed
| Evidence | What the record shows | Source type |
|---|---|---|
| Tournament paper | Results of the Protein Engineering Tournament, Proteins, November 2025 (PMID 40546234). Authors include Kane H and Yusuf A. This is the only PubMed hit for the company. | Public record |
| Tournament placing | Second place behind TUM Rostlab, per a third party summary of the preprint. The company also appears as an organizer. | Independent |
| Y Combinator | Listed as batch S21 with two founders. | Independent |
| Funding | $125K from Y Combinator (PitchBook, Tracxn). CB Insights says $130K. Getlatka says $0. Two later rounds have no amounts. | Sources differ |
| SEC Form D | None found for Medium Biosciences in the records search. One EDGAR hit for a similarly worded query was Droplet Biosciences, an unrelated Massachusetts company. | Not found |
| Federal grants | The records search returned 25 NSF awards matched on the word Biosciences. None names Medium Biosciences or its founders, so none is used here. | Not found |
| Customers and output | 25+ completed projects, 20+ unique proteins engineered, 10+ customers across industry and academia. No customer named, no case study found. | Company-stated |
| Revenue | Getlatka estimates $900K for 2024. The page labels it an estimate and the company has not confirmed it. | Estimate |
| Headquarters | San Francisco per YC, PitchBook and Getlatka. Boston per CB Insights, LifeScienceHistory and Employbl. | Sources differ |
Read plainly: there is real scientific participation, a peer reviewed paper with both founders as authors and a credible accelerator pedigree. There is no independent confirmation of the reagents themselves, no named customer and no reported financing beyond a $125K accelerator check. The sources also disagree on where the company is and what it sells.
What to watch
- Named customers or a published case study showing a designed reagent performing against a commercial antibody.
- Any Form D or priced round. The known investors, AIX Ventures, E14 Fund, Village Global and SBXi, appear on databases without amounts.
- Whether the founders and the company keep a clear line between the tournament they help run and their own entries in it.
- A consistent headquarters, title and product description across the company’s public profiles.
In their words
“We turn advances in AI and protein design into experimentally validated molecular tools, partnering with research teams and diagnostic developers to deliver high-performance affinity reagents in weeks, not months.”
Medium Biosciences, company description, 2026 · Company
“MediumBio and Marks Lab finished second and third, respectively.”
Neermita Bhattacharya, CBIRT, 2024 · Independent
“Entrants came from academia, industry, and research, but they all had only one wish in common: to reach the farthest point of what can be achieved with protein engineering.”
Neermita Bhattacharya, CBIRT, 2024 · Independent
“Our models can be used to suggest high performing sequences that would have typically taken months of wet lab work to identify.”
Medium Biosciences, earlier company description, VentureRadar, 2026 · Company
“We detail the structure and outcomes of a pilot Tournament involving seven protein design teams, powered by six multi-objective datasets, with experimental characterization by our partner, International Flavors and Fragrances.”
Mohamed Hassan Kane, publication summary, LinkedIn, 2023 · Company
“The grand challenge of protein engineering is the development of computational models that can characterize and generate protein sequences for any arbitrary function.”
Tournament authors, abstract, Zenodo, 2023 · Independent
Related companies
Sources
- Public recordResults of the Protein Engineering Tournament, PMID 40546234
- Public recordSEC EDGAR Form D search and records match for Medium Biosciences, no filing found
- Public recordNSF award records match, no award tied to the company
- IndependentInnovating Through Collaboration: The Protein Engineering Tournament’s Impact on Computational Design
- IndependentThe Protein Engineering Tournament: An Open Science Benchmark for Protein Modeling and Design
- IndependentBenchmarks, tournament program page
- IndependentMedium Biosciences 2026 company profile
- IndependentMedium Biosciences funding rounds and investors
- IndependentMedium Biosciences products, financials and employees
- IndependentMedium Biosciences revenue estimate
- IndependentMedium Biosciences YC company profile
- IndependentMedium Biosciences company profile
- IndependentMedium Biosciences profile, jobs and funding
- IndependentMedium Biosciences organisation profile
- CompanyMohamed Hassan Kane profile: publications, patent and projects
- CompanyMedium Biosciences: AI-designed Affinity Reagents
- CompanyCustom affinity reagents, designed for your target and application
Profile researched and written by Healthcare Discovery. Last updated September 29, 2026.
