Final-Year Computer Engineering Undergraduate, University of Ruhuna.
Building production-ready AI systems
from models to scalable software.
Software Engineer • AI/ML • Agentic AI • RAG
AI & ML projects deployed end-to-end (training → API → UI)
Experience with real-time systems, pipelines & model persistence
Research-focused: Blockchain + GNNs for vehicular networks
Product Showcase
Technical Arsenal
Python
Core Language
LangChain
LLM Orchestration
PyTorch
Deep Learning
OpenAI / LLMs
AI Models
RAG Architecture
System Design
HuggingFace
Transformers
TypeScript
Language
Next.js & React
Frontend Core
Node.js & Express
Backend APIs
Vector DBs
Data Storage
Data Analysis
Pandas / Scikit
MLOps / Git
Deployment
Work Experience
Software Engineering Intern
Capricon Solution Pvt Ltd • 6 Months
Developed and maintained enterprise-level solutions, specifically focusing on a robust Point of Sale (POS) system. Handled backend logic for inventory management, sales tracking, and reporting modules to support business operations.
Business Logic
Translated complex business requirements into clean, maintainable code within the MVC architecture.
Database Design
Managed relational databases (MySQL) for high-transaction environments (sales/inventory data).
Featured Projects
View all projectsShowing 4 of 8
NeuroSpace
AI / MLA multi-modal GraphRAG agent that converts video lectures and research papers into a navigable knowledge graph with agentic retrieval and cross-modal navigation.
Key learning: Designed cross-modal GraphRAG pipelines with multi-source ingestion and retrieval orchestration.
ContextIQ (RAG System)
AI / MLChat with your latest PDF using Gemini plus Pinecone RAG. Upload, chunk, embed, retrieve, and answer with grounded context.
Key learning: Implemented reliable RAG flow with robust document chunking and grounded response generation.
AI Data Analyst
AI / MLIntelligent CSV analysis tool powered by Groq and LangChain agents. Upload data and interact using natural language queries.
Key learning: Built agent-powered tabular analysis workflows for natural language data interrogation.
Resume Optimizer
AI / MLA Streamlit dashboard that parses resumes into structured JSON and provides an ATS score using Llama 3.
Key learning: Built structured extraction pipelines with prompt design for ATS-focused feedback.
Engineering Journal & Open Source
Current Focus
from langchain.agents import AgentExecutor
async def run_inference(payload):
# Orchestrating LLM workflow
engine = AgentExecutor(
llm=model,
tools=[rag_search, web_fetch],
verbose=True
)
✓ System ready. Latency: 42ms
Building
Production-grade AI apps with clean APIs, observability, and predictable outputs.
Learning
LLM system prompts, structured JSON extraction, and evaluation for reliability.
Open To
AI/ML engineering, backend systems, and impactful product collaborations.
Open to Conversations
Always open to discussing AI systems, backend scaling, or new opportunities.