Computer Science & Engineering Undergraduate
I build across the stack — software, silicon, and everything between.
I'm Hiruna Malavipathirana, an Integrated Computer Engineering undergraduate focused on software engineering, embedded systems, and robotics. I like problems that span a full system — from firmware talking to a sensor, up through the network, to the interface someone actually uses.
Focus
What I work on
Software Engineering
Full-stack web and mobile apps, APIs, and backend systems — built with an eye toward reliability and clean architecture.
AI & Agentic Systems
Multi-agent LLM copilots, tool-use pipelines, and applied ML — from production agentic systems to computer-vision and trading pipelines.
Embedded Systems & Robotics
Bare-metal and RTOS firmware, sensor fusion, and SLAM — for microcontrollers, autonomous robots, and real-time constraints.
Selected Work
Featured projects
A mix of firmware, systems, and web work — see the full list for more.
Industrial Telemetry & Agentic AI System
Award-winning enterprise tracking platform fusing NFC/GPS/MQTT hardware telemetry with a multi-agent AI copilot for natural-language queries over live operations data.
LoRa Industrial IoT Yield Tracker
Independent LoRa IoT system tracking latex yield across a 2,200+ unit plantation, with custom bucket-level firmware and canopy-penetrating radio links.
Autonomous Inventory Robot (Botzilla)
Sensor-fusion SLAM pipeline for an Nvidia Jetson-powered robot, fusing LiDAR and RGB-D depth to autonomously map and collect cubes.
Autonomous Micromouse Navigation Algorithm
Flood-fill maze-solving and fast-run pathfinding in bare-metal C, deployed to an STM32 board for SLIIT Robofest 2025.
BLNK — AI Design Generator & Co-Designer
Text-to-design AI generator and conversational co-designer for print-on-demand apparel, built on Google Gemini with a self-critique art-director pass.
CAT — AI-Powered CSE Trading Suggestion System
Daily buy/sell/hold suggestions for the Colombo Stock Exchange, fusing an XGBoost model with FinBERT sentiment and LLM-extracted fundamentals.