Dr Peter McCann Strain

Conversico is an AI voice receptionist for UK dental practices. It is in production and in pilot across practices. Greeta answers missed, overflow and out-of-hours calls, supports booking within a practice's rules and hands difficult calls back to a person. I built the website, Greeta application, operator dashboard, architecture, demo and product code. The clinical founders shape the practice workflow and requirements. A multi-agent approach is in development to coordinate several processes inside dental practices.

Pharmaceutical patent research has to connect legal text with chemical structures in patent drawings. I conceived, designed and built Praviar alone to support freedom-to-operate research. It combines AI-assisted research with software checks and human review. Its computer-vision pipeline reads compounds and generic Markush structures from patent drawings and was tested on held-out examples. The product is in production and preparing to go to market.

I joined SThree as one of two founding AI engineers. Since then, I have helped conceive and build its live recruitment platform, which matches candidates to roles, analyses CVs, conducts phone interviews and produces shortlists. I also conceived, designed, built, tested and productionised Aurum, the 25-agent synthetic-data system used to test and improve that platform. The AI engineering team grew from 2 to 20, while the dedicated recruitment-platform team is now close to 100 people.

Datamise supports equal-pay claims involving Birmingham City Council and more than 40 other UK employers, including work in Leeds. Its use continues to grow across UK campaigns. Datamise supported campaigns in which more than £300m in claims were resolved. GMB, legal teams, unions and claimants led the campaigns. I co-founded the company, and Datamise built the wider claims system for GMB's equal-pay work. I personally built SettleMise from scratch as a Mac desktop application and a web application. Both versions were used directly in the Birmingham settlement negotiations. Teams from GMB and Leigh Day could change assumptions and see numbers and charts update while options were discussed. SettleMise continues to be used across equal-pay claims beyond Birmingham. Its use continues to grow.

I built Personal Assistant AI because I wanted to reduce the mental overhead of capturing ideas, remembering context, reviewing commitments, planning and taking action. The working system uses 8 specialist agents. I conceived, designed and built it alone, and I use it every day. The source is now available as the Personal Execution OS public preview. My deployment and data remain private. It is not a hosted product and is not available to buy.

I was the only solo entrant selected, alongside 24 companies, for the £5,000 CivTech 11.5 Exploration grant. I produced an accelerator plan for a human-led AI system covering Scotland's eight impact-assessment types. The work concluded at the Exploration stage and was not deployed.

I spent 2.5 years at Astroscale building satellite position and orientation software specifically for ELSA-M. The pose system reached about four degrees of orientation error. I also built AstroGAN, which substantially reduced the measured gaps on proxy measures, but I left before completing downstream pose-estimation validation.

Completed a DPhil in AI and Clinical Medicine, funded by a £125,000 Medical Research Council scholarship. The research used deep learning to study mouse placental histology, including cell detection and classification, whole-slide tissue segmentation and analysis of cell morphology across genotypes and phenotypes.