Co-developed an AI email management platform with one teammate. Personally researched and built voice dictation, inline email summaries, advanced Gmail-style filters, and full Google Calendar integration to client requirements.
I build software people actually use.
Full-stack engineer building AI-powered products, scalable backends, and modern frontends. React, Next.js, Node.js, Python, and LLMs — from prototype to production.

Products that show how I build full-stack AI products.
→ Hover a folder to pull it open · click ▸ Open file for the full case study
EduFlow, built end to end for client testing, and Draftlee, a two-person project — alongside my open-source production systems.






Full-stack contributor · Two-person team
Collaborative client project
Co-developed an AI email management platform with one teammate. Personally researched and built voice dictation, inline email summaries, advanced Gmail-style filters, and full Google Calendar integration to client requirements.

Sole developer · End-to-end full-stack & AI
In development · Deploying for client testing
Built the complete AI-powered LMS for mortgage advisors and paraplanners: secure course delivery, contextual voice Q&A, AI-generated assessments, intelligent tutoring, CPD tracking, and team analytics. Deploying for client testing during ongoing development.

Clinical document intelligence for medical reports: upload lab results and prescriptions, extract structured facts with page-level provenance, run deterministic safety checks, then ask questions that cite the source page.

Multi-tenant SaaS for mushroom farms — cultivation batches, environmental records, inventory, purchasing, harvest grading, analytics, AI assistance and Stripe billing, with isolation enforced by both role and organization.

Medical document intelligence in an anonymous workspace — no signup, just a session id in the browser. Uploads become a timeline of visits, medications, labs and allergies, with deterministic trend checks and grounded answers that cite their source.

Transit platform with real-time vehicle tracking and seat reservation — drivers broadcast GPS over WebSockets, riders watch the bus move and book from a live 32-seat grid, with Stripe checkout and an offline PWA ticket vault.
One engineer, connected across the stack.
Hover a product window or tap a capability below to trace how AI pipelines, backend APIs, and modern frontends connect across what I've shipped.









Built for speed, scale & intelligence.
A focused toolkit for shipping from prototype to production.
Frontend
Component systems, performance, and product-grade UX
Backend
APIs, services, and scalable server logic
AI / Machine Learning
LLMs, RAG, embeddings, and AI pipelines
Databases
Data modeling and persistence
Cloud / DevOps
Deployment and delivery
Where I've made measurable impact.
AI platforms, automation tools, and full-stack systems in production.
EduFlow — AI-Powered Learning Platform
- ▪Extracted PDF and PowerPoint content with multiple fallbacks, including Tesseract.js OCR for image-based PDFs and pdf2pic/sharp image processing.
- ▪Built screen-aware voice Q&A using full-document and page/slide-specific context, transcription, fuzzy matching, and spoken responses.
- ▪Generated adaptive MCQs from document chunks, with quality controls and new question pools for retakes.
- ▪Synchronized PDF-page, video-time, and slide progress with CPD hours, certificates, dashboards, and manager notifications.
Draftlee — AI Email Management Platform
- ▪Researched and developed voice input for email dictation.
- ▪Built inline email summaries within the inbox workflow.
- ▪Implemented detailed Gmail-style email filters.
- ▪Developed the full Google Calendar integration to client requirements, including availability, event creation, invitations, time zones, and conflict detection.
Learning that compounds over time.
Systems design, algorithms, and software engineering — paired with production shipping and IEEE research.
BSc (Hons) in Information Technology
Systems design, algorithms, and software engineering — with hands-on AI research and production shipping. IEEE ICAC 2024 publication.
IEEE Publication
Peer-reviewed research on educational technology and student well-being.
Analyzing Academic, Relational, Economic, and Social Influences on University Student Happiness
A Python web-based machine learning and artificial intelligence research project analyzing academic, relational, economic, and social influences on university student happiness. This study leverages AI models and data analytics to provide insights into educational technology and well-being.
Turning ideas into products.
National recognition, seed funding, and a first stage speech.
SPARK 101
Top 101 Young Entrepreneurial Talents — Sri Lanka
Selected among Sri Lanka's top young entrepreneurial talents at the SPARK Grand Finale, powered by the Ceylon Chamber of Commerce, ILO, and U.S. Embassy.
Thalir Seed-Funding
Top 4 Startup Idea Champions
First-ever funding pitch and stage speech. Digital product idea validated with a seed grant from David Pieris Group's Thalir Program.
Continuous technical development.
Certifications, research, and learning milestones mapped to real production work.
Advanced React & Frontend Architecture
Next.js - Full Stack Development
Node.js Backend & REST APIs
OpenAI - Building AI-Powered Applications
RAG Systems - Retrieval Augmented Generation
IEEE ICAC Research Publication
Engineer who ships AI that people actually use.
Software Engineer building production AI products across the full stack — from AI-powered learning platforms and email automation to document processing pipelines. Core stack: React, Next.js, Node.js, Python, OpenAI, Gemini.
Available for freelance and full-time roles — building from scratch or strengthening existing teams.
Let's build something great.
Tell me about the product you're building.