CommPAL

Back to Activities

Professional Consulting/Initiatives

Technologies

AWS Docker TensorFlow OR-Tools Gurobi SAM Terraform React +3 more

June 2022 - Present

University College, Cork (Remote)

Gallery

Technical Lead

AI-based digital health startup moving from research to commercialization

CommPAL is an AI-based digital health startup. Development began following several years of research. I was hired to interpret the research, develop an MVP, and move the project into commercialisation phase.

Challenge

CommPAL needed to transform years of academic research into a commercially viable digital health platform while meeting strict medical device regulations and ensuring scalable ML infrastructure.

Key Points

Translating academic research into practical technology
Building a compliant system that meets ISO-62366 and ISO-62304 standards
Creating a scalable architecture for ML models and data
Managing complex stakeholder requirements

The Approach

Discovery Phase:

Analyzed research findings and requirements
Identified key functional and regulatory requirements
Conducted stakeholder interviews and needs assessment

Planning & Strategy:

Developed microservices architecture for maintainability and compliance
Created implementation roadmap with regulatory milestones
Designed data governance framework for healthcare compliance

Design Considerations:

Privacy-by-design approach to meet GDPR requirements
Service-oriented architecture for scalability
Compliance-first feature implementation

Achievements

📈 Project Management

Setup, management, and organisation of initial projects
Managed technical team & externals
Implemented Sharepoint, Notion & JIRA

☁️ Model as a Service & SaaS

Developed platforms using AWS Services
Implemented ML-based micro-services architecture with Docker, CPLEX, and OR-Tools
Managed A-Z deployment using SAM & Terraform

🌐 Web Development

Built web applications with React, Vue and Next.js
Developed & managed websites using Tailwind and React
Implemented SSO & other forms of authentication

🤖 Machine Learning

Deployed and managed TensorFlow, OR-Tools, and Gurobi using Fargate, EC2
Liaised with ML engineers to develop optimal stack, architecture, and feature sets
Built data lake (w/ medallion layering) and data engineering pipelines
Implemented model testing & versioning

🔒 Regulatory and Compliance

Managed Data Protection Impact Assessments (DPIA), MDR compliance, handover documentation, and team training
Oversaw security and compliance implementations aligning with ISO-6266, ISO-62304, privacy regulations, QMS contributions, specification documents, and pentesting

Lessons Learned

What Went Well:

Microservices architecture provided flexibility for evolving requirements
Early compliance integration saved time in later development stages
ML versioning system ensured reproducible results and regulatory traceability

Key Takeaways:

Healthcare AI requires specialized compliance knowledge beyond typical software
Research translation benefits from iterative development with researcher feedback
Data governance is as critical as the ML models themselves