Agentic assistants, knowledge graphs, RAG, and LLM observability across CloudofGoods, About.lk and Xenia. → see the projects
:~$cat whoami.html
About me
associate ai engineer · battaramulla, sri lanka
I build AI features for real applications. The part I find most interesting is autonomy — how an agent decides, and how to make a system that reasons, retrieves, and acts without a human in every step. I try to stay tech-agnostic and work from first principles: understand the problem, find the real bottleneck, then build.
I earned my BSc (Hons) in Data Science from SLTC Research University, with early industrial exposure as an AI/Data Engineer Intern at Michelin Lanka. Today I work as an Associate AI Engineer at iLabs, where I've architected agentic assistants, knowledge-graph pipelines, and LLM observability stacks. Before all that — and still on the side — I spent years freelancing, shipping 50+ web, mobile and IoT apps.
what i'm into
- Agentic systems — orchestrators, sub-agents, tools, and making them reliable.
- Knowledge graphs + RAG — giving models structure and memory, not just a context window.
- LLM observability — being able to see what a model is actually doing.
the stack
tools i reach for, grouped by what they're for.
the timeline
Predictive bottleneck modeling (SimPy + XGBoost) and an automated ETL pipeline that cut reporting from hours to seconds.
50+ shipped apps across web, mobile and IoT — computer vision, deep learning, ESP32 embedded systems, full-stack across e-commerce, tourism and LMS.
Final-year research on an Airflow-orchestrated data platform + a Transformer for sequential price prediction. → research details