CAISY - Clever Artificial Intelligence System
CAISY is an interactive patient-feedback robot and companion mobile app that replaces traditional pen-and-paper surveys. It’s backed by a custom SaaS analytics platform that aggregates responses for powerful reporting across clinics and departments. Designed for children, teens, and adults, CAISY is used in hospital outpatient services and can be accessed at home for contactless, COVID-safe feedback collection.
Machine learning
CAISY is powered by TensorFlow.js. I combined several pre-trained models to give the robot awareness of its environment, including real‑time hand detection for touchless interaction. It is intended for CAISY to understand sign language, including British Sign Language (BSL) and ASL.
Robotics
CAISY is a static, kiosk‑style interactive robot designed for reliability in clinical environments. The hardware and enclosure support approachable, low‑friction interactions while guiding users through short, age‑appropriate questionnaires.
Prototyping
I iterated through multiple physical and UI prototypes to validate interaction patterns, accessibility, and engagement across age groups. This included paper mockups, 3D form studies, and working electronics to test durability and clinic‑flow fit.
Web App & SaaS
At the heart of CAISY is a custom‑built SaaS analytics backend for secure data aggregation and reporting. Each deployed unit uploads anonymised responses to a central dashboard where staff can filter by site, timeframe, or questionnaire, export CSVs, and track trends over time.
Partnership with UK NHS Hospitals
Secured early‑stage NHS funding (£3.5k). Partnered with AHSN (Academic Health Science Network) and NHS Trust (NHS CDDFT) to pilot and license the robot and SaaS with hospitals and GP practices across the UK. Winner of the Teesside LaunchPad Business Start‑up Programme (2021) with regional press coverage.
Like this project?
I design and build end‑to‑end products across ML, hardware, and web. If you’d like to discuss a project, get in touch.