Mithrl, which builds AI software for pharmaceutical and biotechnology research teams, has disclosed a $20 million Series A. Its Scientific Decision Engine lets scientists request analyses in everyday language and explore what their experimental results mean, reducing the coding work between a laboratory experiment and the next research decision.
The company confirms the round on its website, alongside a September 15 announcement describing its next stage of development. Mithrl previously announced a $4 million seed round led by Bonfire Ventures in November 2024, when it outlined plans to expand its commercial team and develop the platform.
A scientist working with sequencing data can ask the software to run an analysis, then investigate the findings against other datasets and scientific knowledge. Mithrl's platform page says researchers can do this without coding and describes the resulting work as traceable and reproducible. The practical aim is to let the people designing experiments explore their results directly, with less dependence on a specialist to build each computational workflow.
The enterprise offering combines software with scientific and engineering support. Mithrl says it installs the technology inside partners' environments under their governance, while its staff develop analyses and workflows around individual research programs. That approach accommodates differences in assays, data and standards of evidence across drug developers, but also makes hands-on implementation part of the business.
An April case study describes an unnamed clinical-stage cell-therapy company that used Mithrl to analyze single-cell sequencing data. Mithrl says analysis fell from eight to ten weeks to less than one week, with findings consistent with human bioinformaticians' work. Scientists, project managers and leadership also gained direct access to explore results. Those are outcomes reported for one deployment, rather than an established performance benchmark across customers.
In its September announcement, Mithrl says it plans to launch its Biomedical World Model the following week, connecting experimental results with structured biological knowledge to help researchers compare explanations and decide what to test next. The business question is whether it can carry that evidence through successive stages of a drug program while making each deployment less dependent on bespoke scientific and engineering work.