Celeris Therapeutics and the EIC Accelerator: AI Drugs to Bankruptcy

Most disease-causing proteins cannot be blocked by conventional drugs, and Celeris Therapeutics wanted to have them destroyed instead. Targeted protein degraders tag a harmful protein so the cell's own disposal system breaks it down. Celeris, founded by Christopher Trummer and Jakob Hohenberger, built Celeris One, a deep-learning platform to design such molecules, with a US company in Menlo Park and an Austrian company, registered in Graz in 2021, according to the European Investment Fund. In 2022 it signed discovery collaborations with Merck KGaA and Boehringer Ingelheim.

The European Innovation Council granted €2,500,000 towards a €4.9 million project from July 2022 to February 2024. The final report says the company finished an 800 square metre biosafety level 2 lab in Graz with about 22 full-time staff, and achieved "positive biological validation of one of our novel compounds for our Oncology I program as hit". On 15 October 2024 the Graz regional court appointed an insolvency administrator for Celeris Therapeutics GmbH in bankruptcy proceedings, according to North Data.

What Happened to Celeris

The Austrian startup news site brutkasten reported the bankruptcy of the Austrian subsidiary of the biotech "with Graz roots". The register records a distribution to creditors on 17 July 2026. The company's news page returned an error when checked in September 2026. We found no announcement from the US parent about the future of the platform, its drug programmes or its pharma collaborations.

The Technology: Predicting Three-Way Complexes

A degrader works only if it brings the target protein and a cellular disposal enzyme together in a stable three-part complex. The EIC report describes "novel graph-based algorithms to predict the formation of ternary complexes", in-house tissue culture, microbiology and imaging, and three oncology programmes. The one validated hit showed protein degradation in western blots and binding data, according to the report. No compound entered clinical development.

What the EIC Project Promised, and What CORDIS Shows

The project was to build a closed-loop, cloud-based deep learning platform for early drug discovery in protein degradation. It built the platform, the lab and a first hit, and published papers. It did not produce a drug candidate, and the company that ran it failed within months of the grant ending.

1. Wet labBuilt800 m², Graz.
2. Validated compoundOne hitOncology I.
3. Drug candidateNoneNo clinical entry.
4. Company survivalFailedBankrupt Oct 2024.

The Money

The Austrian company's register lists €600,000 in seed financing support in 2021, €822,752 in FFG state aid in 2022 and the €2.5 million EIC grant. BioSpace reported that Celeris had raised $4.4 million in February 2022. Whether the equity part of its blended EIC award was ever invested is not public. Its last filed accounts cover the year to May 2023.

What Cannot Be Checked

Why Europe Should Care

The EIC funded a promising AI drug discovery start-up that had already won two big pharma partners. Its Austrian company went bankrupt seven months after the grant ended; the public record does not show what became of the work the EU paid for.

The Verdict

Celeris Therapeutics, its insolvency administrator and the EIC have not been asked for comment for this article. It is based entirely on public records and published statements, linked throughout.