€6.3m · Series B · Agritech, Healthtech · Tübingen, Germany
Computomics, a Tübingen company applying machine learning to plant breeding, has raised €6.3m in Series B equity financing. Convent Capital Agri Food Fund led the round with €5m; existing investors High-Tech Gründerfonds, MBG Baden-Württemberg and Amathaon Capital participated alongside founders and scientific advisers. The company says the capital will scale commercial delivery of its climate-smart breeding platform.
Turning field trials into earlier breeding decisions
Plant breeders must choose which crosses and candidate varieties to advance years before the final crop reaches the market. Field testing remains essential, but the number of possible genetic combinations is far larger than any programme can trial. Computomics’ ×SeedScore platform combines genotypic and phenotypic records with environmental data such as temperature, rainfall and soil conditions, then predicts how candidates may perform across locations and future climates. The output is designed to help breeding teams rank crosses, focus field capacity and decide where varieties should be tested or placed.
The commercial mechanism depends on fitting the technology to each customer’s breeding programme rather than offering one generic model. Computomics says customer datasets remain separate, models are validated against unseen data and predictions are refreshed as new seasons of results arrive. That recurring assess-train-predict-select cycle can embed the company in long-running research workflows, but it also makes data quality, biological validation and consistent project delivery central to customer value. The financing therefore supports a delivery challenge as much as a modelling one: expanding to more breeding programmes without weakening the programme-specific work that makes the predictions useful.
Proof points from commercial breeding
The company was founded in 2012 and works with commercial breeders across field crops, forage, vegetables and specialty crops. In customer case studies, AB InBev describes using ×SeedScore to model barley crosses against genetics, field measurements, climate and drone data, while Beck’s says it uses the platform to narrow hundreds of thousands of possible maize crosses before field trials. These are company-published customer claims, but they illustrate where the product is meant to earn its place: reducing low-value testing while preserving field evidence for the most promising candidates.
Convent’s investment thesis links commercial adoption to environmental impact: if better predictions help breeders avoid wasted seasons and develop varieties suited to heat or drought, the climate benefit should grow with platform use. The immediate execution test is narrower and measurable: whether Computomics can turn bespoke data-science engagements into a repeatable delivery model across more crops, customers and geographies.
