TileBio LTD
TileBio builds a self-learning AI model for pathology that reads whole-slide pathology images at scale without human labels, giving pharma and biotech teams preclinical toxicity and tissue-abnormality detection, archive search by tissue morphology, and exploratory biomarker discovery and is developing a clinical triage tool for use within the NHS.
DiagnosticsOn the market (pre-revenue)United Kingdom
See the full profile
The pitch deck, traction, financials and competitive detail are visible to approved investors, mentors and corporates. Log in or join EndCancer.Life to request access.
Company info
- Countries of Operation
- United Kingdom
- Founding Year
- 2025
- Number of Employees
- 6–10
Company overview
- Full Description
- TileBio is a Glasgow-based computational pathology company founded in 2025 by Chris Walsh (CEO), Ke Yuan (CSO), John Le Quesne (CMO Pathology) and David Chang (CMO Medical). The company builds a self-learning AI that reads whole-slide histology images without labels, turning tissue morphology into a discrete vocabulary that can be searched, quantified and modelled at scale. Its tissue-language model has been trained on around 100,000 whole-slide images, with training on more than 4 million slides from NHS Greater Glasgow and Clyde planned by mid-2027. TileBio holds exclusive commercial rights to approximately 4 million NHS Scotland pathology slides, with outcome linkage on a subset of around 100,000 cases. The core technology was developed at the University of Glasgow, is patent protected and exclusively licensed to TileBio. The company sells into pharma and biotech with two initial focuses: preclinical toxicologic pathology (tissue abnormality detection and morphology-based search of archival tox slides) and exploratory biomarker discovery for oncology clinical trials. TileBio is also working towards a clinical triage tool for use within the NHS. The company closed a £2.3M seed round in May 2026 backed by TwinPath Ventures, Scottish Enterprise, Maven Capital and GU Holdings, and employs around ten people.
- Areas of focus
- Computational pathology / AI/ML foundation models / Preclinical toxicology / Biomarker discovery
- Problem Statement
- Tissue morphology is the richest readout in oncology and drug development, yet it remains largely locked away. Pathologists read slides one at a time, scoring is semi-quantitative and inconsistent between readers, and the hundreds of millions of archived slides held by hospitals and pharma companies cannot be searched by what the tissue actually looks like. In preclinical toxicology, pathologists spend weeks investigating unexpected safety findings with no way to ask "have we seen this before?" across a company's archive. In clinical trials, histology endpoints are coarse, so real biological signal is lost and promising programmes fail or stall. Existing AI tools depend on expensive expert annotation, target narrow tasks and do not scale across tissues, species and indications. The result is slower, costlier drug development and cancer diagnostics that leave most of the information in the tissue unused.
Interested in this innovator?
Send a short message. We’ll pass it to TileBio LTD with your name and email, so they can reply.
Request an introduction
Create an account or log in to send this request. Already a member? Log in
Introductions to innovators are open to innovators, investors, mentors and corporates.