Orakl Oncology raises $34M Series A to predict cancer drug behaviour before the clinical trial

Orakl Oncology (Paris) has raised $34 million (approximately €31 million) in a Series A co-led by Norrsken VC and Singular, with InfraVia, Serena, Bpifrance (Patient Autonome Fund), teampact.ventures, Plug and Play, Verve Ventures, Speedinvest, and returning investor HCVC participating. Angels include Thomas Wolf (Hugging Face co-founder) and Stan Polu. The company, spun out of Institut Gustave Roussy in 2023, will use the capital to expand its lab capacity fivefold, extend into lung and gastric cancer, and open commercial teams in the US and UK.

The failure-rate problem the industry has learned to live with

Roughly 90 per cent of oncology drugs that enter clinical trials fail to reach approval. The standard model — test compounds in cell lines and mice, run a Phase I to establish safety, then a Phase II and Phase III to find out whether the drug works in humans — has remained structurally unchanged for decades. Its failure rate is not a secret; it is a cost that the industry builds into its economics. A drug that costs $2 billion and 15 years to develop and then fails in Phase III has usually been failing, quietly, since preclinical validation.

Orakl Oncology's premise is that this failure rate is not inevitable — it is a measurement problem. The company builds patient-derived tumour organoids: three-dimensional tissue models grown from real patient samples that preserve the cellular architecture, genetic heterogeneity, and micro-environment of the original tumour far more faithfully than flat cell cultures or animal models. It then layers an AI platform on top to find the patterns in organoid phenotype data that predict clinical response.

The company was spun out of Institut Gustave Roussy — Europe's largest cancer centre, located outside Paris — giving it access to the patient sample base and clinical expertise that most biotech startups have to build from scratch over years.

Singular's second bet this week

The co-lead structure is notable. Norrsken VC, the impact fund backed by Klarna co-founder Niklas Adalberth, invests across health, climate, and financial inclusion — with a track record in Healthtech that has included Eyedentity from this archive. Singular, the Paris-based fund known for early bets on AI-native vertical software, backed Sumary two days ago with a $25M seed in Copenhagen: an AI platform that automates the volume work of accounting. The two deals share a common structure — domain expertise meets AI to compress a cycle that the industry had accepted as a fixed cost — but the thesis is applied to cancer drug development at a scale and risk profile an order of magnitude higher.

Thomas Wolf among the angels is a significant signal: the Hugging Face co-founder's bet on an oncology platform speaks to confidence in the AI-biology convergence, not in AI tooling generically.

Fivefold capacity expansion and the data bet

The $34M goes primarily into scaling lab capacity — described as a fivefold increase in data generation — and into extending the platform beyond its initial cancer type into lung and gastric cancer. Each new indication requires new organoid models, new training data, and validation against real patient outcomes. The US and UK commercial teams are the revenue infrastructure; the lab expansion is the scientific one.

The core question at this stage of Orakl's growth is whether the platform can generate data fast enough that the AI predictions become generalisable across patient populations with different genetic backgrounds — and whether that validation happens before pharma customers require it.