AI Systems Engineer | RAG & Agentic AI | Full-Stack ML Developer
I am a final-year Computer Engineering student at the Faculty of Engineering, University of Ruhuna, specializing in Artificial Intelligence, Machine Learning, and Full-Stack development. I design and deploy AI systems that move beyond notebooks into production. My work focuses on Retrieval-Augmented Generation (RAG), multi-modal knowledge graphs, and scalable ML pipelines.
- 🔭 Current Focus: Building Agentic RAG systems and Multi-Modal Knowledge Graphs (GraphRAG).
- 🎓 Education: B.Sc. Engineering (Honours) in Computer Engineering.
- 👯 Looking to collaborate on: AI/ML research, Computer Vision projects, and Open Source.
- 💬 Ask me about: AI/ML, React Native, Node.js, Computer Vision
- Multi-Modal GraphRAG Systems (NeuroSpace)
- Agentic AI workflows with tool-calling
- Scalable MLOps pipelines with Docker + CI/CD
- Research on Blockchain + Graph Neural Networks(TGNNS) for SDVNs
Languages
AI & Machine Learning
Generative AI & Agentic RAG
Web Development
Tools & Cloud
- 6+ Production Deployed Applications
- Built end-to-end RAG systems using Pinecone & LangChain
- Designed ML pipelines with CI/CD + Docker
- Real-time systems using WebSockets
| Project | Live Demo | Description | Tech Stack | Status |
|---|---|---|---|---|
| NeuroSpace | 🔗 Coming Soon | A Multi-Modal GraphRAG Agent that converts video lectures & research papers into a navigable knowledge graph with agentic retrieval and cross-modal navigation. | LlamaIndex, Neo4j, React Flow, FastAPI, Docker |  |
| ContextIQ (RAG System) |  |
Chat with your latest PDF using Gemini + Pinecone RAG. Upload → chunk → embed → retrieve → answer with grounded context. | Python, Streamlit, Gemini, Pinecone, LangChain |  |
| AI Data Analyst | |
Intelligent CSV analysis tool powered by Groq + LangChain agents. Upload data and interact using natural language queries. | Python, Streamlit, LangChain, Groq, Pandas | |
| Salary Estimator (MLOps Pipeline) | |
End-to-end ML system with CI/CD, Docker containerization, automated testing, and production deployment. | Python, Flask, Docker, GitHub Actions, Render | |
| EV Charging Booking System | |
Real-time EV charging booking & tracking system featuring route optimization and live station monitoring. | React Native, Node.js, Socket.io | |
| Live Talk (Encrypted Chat App) | |
End-to-end encrypted real-time communication platform with secure file transfer and multi-user channels. | React, Redis, Socket.io | |
| Voxis | 🔗 Coming Soon | A voice assistant named Jarvis that records audio, transcribes speech using Whisper, and generates conversational responses with Llama AI. | Python, Groq, Whisper, sounddevice, NumPy, SciPy | |
- Graph Neural Networks (GNN)
- Blockchain Systems for Vehicular Networks
- Multi-Agent AI Systems
- Knowledge Graph Construction
- Mathematics Teacher
- Technical Presenter & Public Speaker
- AI Systems Enthusiast
- Portfolio: https://diluksha-upeka.me
- LinkedIn: (https://www.linkedin.com/in/upeka-diluksha/)
- Email: (dilukshaupeka@gmail.com)