AI Clinical Skin Assessment Platform

A UK-founded telehealth startup set out to make clinical-grade skin assessment available without a clinic visit. Patients photograph and describe a skin concern; the platform structures the intake, runs AI-assisted analysis, and routes the case to a clinician for review. The founding team had clinical expertise and early traction, and needed an engineering partner who could build the production system around the AI: PHI-conscious data flows, an audit trail for every AI-assisted step, and a review workflow clinicians would trust.

A 16-week fixed-scope build covering the full loop: guided patient intake designed to keep completion high, an AI-assisted assessment pipeline with the model kept strictly assistive, a provider review queue with full case context, and an append-only audit trail recording what the model saw and suggested for every case. The architecture treats HIPAA as a design input rather than a launch checklist: PHI lives only in services covered by the right agreements, and nothing the model touches goes unlogged.

Clinicians review cases faster because intake arrives structured and pre-assessed, and patients finish the intake at rates consumer health products rarely see. The working product went on to feature in the client’s seed fundraising materials. The engagement is the template for how we approach clinical AI: the model assists, the clinician decides, and the audit trail proves both.