ECG & PPG Signal Processing SaaS Dashboard
ecg upload, tested filters, r-peaks and rules, plus dashboards
Health Tech
ECG Processing
Role
AI Engineer
Timeline
October 2025 - March 2026
team
Torabisignals & Leeds Trinity University
platform
Web

The Problem
The project is a multi-repository prototype for accepting ECG files, processing the signals, presenting reports and administering accounts and subscriptions. it was produced for a health-technology software project, and the source files describe it as a prototype and web mvp, not a finished product. the web path covers upload, analysis, reports and two dashboards, one for users and one for administrators.
The project also included a placement-year collaboration with Leeds Trinity University. i had the chance to supervise 10-15 placement-year students, double-checking the assignments and plans they had to meet for Python developer and web developer positions.

Constraints
The main constraint is that this is a prototype and web MVP, with material still untested, stubbed or only planned. no trained model artefacts are present, so the learned classification path cannot run and detection falls back to the rules. scope was narrowed as well: the mvp release covers web only, a documented rust gateway was deactivated in favour of python, and the wearable and mobile codebases sit outside that release path. verification is bounded the same way: the files do not establish clinical validation, regulatory approval or a production operating history.



Outcome
The web MVP covers ECG file upload, filtering, R-peak detection, signal-quality checks, heart rate and HRV, and rule-based arrhythmia detection. The tested parts are filtering, R-peak detection, heart rate and HRV, and the arrhythmia rules; the rest is implemented but untested. The web MVP also includes a Next.js user dashboard, a Vite admin dashboard, generated reports, account and subscription administration, and document and structured-data export paths.

What I'd Change
The evidence-led changes match what the file history supports: consolidate the schema migrations and authentication contracts, replace the process-local security state, add a CI pipeline, and validate the signal processing against reference data.
You can see a sample of the project code here
You can see a sample of the project code here
