Projects

Lightweight AI Chrome Extension for summarising, explaining, and re-phrasing highlighted text directly in-browser.

• •Built custom side-panel UI with markdown-formatted AI responses.
• •Integrated context-menu actions for summarise, tone-change, explain, and key-point extraction.

Property recommendations using Gemini+Daft APIs with prompt engineering and interactive UI.

• •Secured 1st place in Google AI Hackathon Dublin 2025.
• •Implemented data ingestion & LLM prompts for curated suggestions.

Event-driven engine for real-time user behaviour analysis and personalised recommendations.

• •Built Kafka-based pipeline for capturing live user events.
• •Implemented persona clustering for real-time recommendations.

AI-assisted triage system enabling real-time patient intake, clinician decisions, and appointment updates.

• •Built contract-driven monorepo with Next.js dashboard, Expo mobile app, and shared TypeScript packages.
• •Implemented end-to-end triage flow from patient survey → backend analysis → clinician approval → appointment update.

Backend lead for an image-classification web app used to refine recycling AI models.

• •Designed & implemented upload pipeline to Azure Blob Storage; converted images to signed URIs for ML ingestion.
• •Optimised portal throughput and latency; improved labeling workflow.

Python-based tool for sending personalised bulk emails using dynamic templates and contact data sources.

• •Automated delivery of 60+ personalised emails from structured contact data.
• •Switched from Excel to JSON input for greater flexibility and cleaner data handling.