blackthorn.ai

Teams building medical devices that run on artificial intelligence face a particular hurdle: regulators want evidence, and the evidence has to be designed before it is collected. Blackthorn AI, which trades under the address blackthorn.ai, has made that its current focus. It says its work sits at the meeting point of software engineering, AI, bioinformatics and the design of scientific workflows.

Evidence for FDA and MDR submissions

For AI medical devices, the company writes statistical analysis plans, validates models and prepares change control documentation. It says researchers who publish their methods write this material, so that it stands up to scrutiny from a notified body and from the FDA. Engagements are offered at fixed scope and fixed price, with a start said to be possible within two weeks.

Data products for pharma and biotech

Alongside regulatory work, the company builds data and AI tools for drug developers:

• OmicsLayer, which takes disorganized biomarker data from several modalities and shapes it into repeatable pharma workflows for choosing cohorts, predicting responders, analysing patients and assembling evidence
• TrialLens, which analyzes patient-level clinical trial data to find responder profiles, non-responder patterns, adverse-event risks and subgroup signals of commercial value
• AssetLens, which estimates development costs, commercial potential and net present value for go-alone and out-licensing scenarios

Project examples

Its case studies include molecule generation for a biotech company developing skin therapies, a digital health platform using vision AI for breast cancer detection, a centralized multi-omics data warehouse that replaced manual data handling with structured pipelines, a genomics platform built on AWS and GATK for a biopharma firm, an AI platform for drug and target discovery, and a voice assistant with a virtual reality health studio for a start-up.

Wider engineering services

Earlier descriptions of the company set out sector work in life and chemical sciences, healthcare, cybersecurity and financial services, with services in AI software development, analytics, data warehousing and MLOps. The technical range includes machine learning, deep learning, computer vision and natural language processing.

How it describes itself

Blackthorn presents its team as AI engineers and scientists, with certified developers holding ten or more years of industry experience. It says it keeps a flat structure so decisions move quickly, and that it aims to be a partner judged on value created for the client rather than an outsourcing firm selling developers' hours.