AI engineer · Researcher · Builder
I build reliable AI systems, invent new tools, and make existing ones better.
01 — About
I like building things that don’t exist yet — and fixing the ones that almost work. Right now, that means making AI systems reliable enough to trust.
Peer-reviewed work on making language models cheaper and more dependable in production — including a first-author paper at IEEE ISPASS.
Salli and Pinglo at Leaf Monkey Labs — AI that does real work for real customers, with deterministic engines wherever the numbers have to be right.
Technical leadership on agentic AI at Gapstars, and teaching AI beyond the theory at the University of Moratuwa.
02 — Now building
At Leaf Monkey Labs, a small studio in Colombo: AI that does real work for real customers, with the deterministic parts done properly.
Stop guessing. Start knowing.
Personal finance and tax planning built for Sri Lanka: a real double-entry ledger, a deterministic IRD tax engine, and an AI advisor that explains your numbers but never makes them up.
Turn every enquiry into a paid booking.
An AI booking assistant for businesses booked by the hour — courts, lessons, coaching. It answers on WhatsApp, Instagram, LINE and web chat, checks the real calendar, holds the slot and takes payment, and hands over to a human any time.
03 — Research
Papers on what it really takes to run AI in production: when small open models can replace big APIs, and how the structure of your code can do the prompt engineering.
04 — Journey
From smart-home research at Moratuwa to technical leadership at Gapstars, founding Leaf Monkey Labs, and teaching back at Moratuwa.
Now
Agentic AI with Stekz, and Data/AI Guild Master — turning ambitious product ideas into dependable, shipped software.
Now
Building Salli, personal finance and tax for Sri Lanka, and Pinglo, an AI booking assistant for businesses booked by the hour.
Now
Back where it started — showing students the world of AI beyond the theory.
2024 — 2025
Virtusa
Agentic customer experience for UnitedHealth Group, built with Google.
2022 — 2024
Jaseci Labs
Research and engineering on LLM integration for the Jac language. First-authored “Scaling Down to Scale Up” (ISPASS ’24).
2017 — 2022
University of Moratuwa
Where the smart-home research started — HomeIO at IEEE AIIoT, later extended in Sensors.
Previously also at promiseQ.
05 — Speaking
I speak about building AI people can depend on — for engineers, founders and students.
What it takes to make LLM systems dependable: evaluation, guardrails, and designing for the day the model is wrong.
Lessons from “Scaling Down to Scale Up” — when self-hosted small language models beat GPT-class APIs, and when they don’t.
What it takes to let an AI agent take bookings and payments for a real business: guardrails, human handoff, and knowing when not to answer.
Meaning-typed programming and prompting: letting the structure of your code do the prompt engineering.
From the blog
06 — Contact · Full moon