Portfolio
Projects
A live, deployed AI assistant alongside published research and applied ML. Open any project for a live demo or an interactive deep dive — architecture, results, and technical detail.
Featured
Generative & Agentic AI · Live demo
BuildRight AI — Guardrailed, Agentic RAG Commerce Assistant
A full-stack hardware-store assistant: an agentic Claude tool-use loop over a 10K+ SKU catalog, hybrid (lexical + vector + visual) RAG with reciprocal-rank fusion and re-ranking, and a deterministic guardrail that validates every price and citation against retrieved data — so a fabricated price can't render. A Haiku→Sonnet router keeps it cheap; it ships as one Docker container.
Hybrid RAG
CLIP visual search
Deterministic guardrails
Haiku→Sonnet routing
Agentic AI
LLM-Powered Multi-Agent Chatbot for Enterprise User Provisioning
A five-agent system (Google ADK, Gemini) automating onboarding/offboarding via a supervisor delegating to policy, identity, and approval agents — with human-in-the-loop gates and a custom MCP server over live Entra ID.
Google ADK
Gemini
MCP
A2A
Explore →
Industry Deployment · Mitacs
Supermarket Refrigerant Leak Detection
ML leak-detection deployed across supermarket refrigeration systems — cut detection latency from months to days across 170+ subsystems per site. Shipped on Azure Functions with CI/CD, automated retraining, and drift monitoring.
XGBoost
PySpark
Azure
MLOps
Explore →
GNN Research
DualSTGF — Dual Spectral-Temporal Graph Fusion
Dual-view spectral-temporal graph fusion with adaptive edge learning and divergence scoring for incipient fault detection. Submitted to Information Fusion (IF 15.5).
PyTorch
PyG
Spectral GNN
Explore →
GNN Research · Published
Dual-Graph GNN for Structural Damage Detection
Dual adjacency matrices (static + feature) with adaptive fusion, 1D-CNN compression, and squeeze-and-excitation for SHM. Published in Engineering Structures (Q1, IF 6.4, 2025).
PyTorch
PyG
SHM
Explore →
Explainable AI
NE-CFE — Node-Edge Counterfactual Explanations
Perturbation-based counterfactuals that jointly optimize node features and edge toggles via Binary-Concrete edge gates, making GNN predictions interpretable for safety-critical applications.
PyTorch
PyG
GNN
Explore →
Explainable AI · IEEE ICAC 2025
Counterfactual Explanations via Conic Optimization
A novel HEOM + Mahalanobis objective for plausible counterfactual generation using quasi-convex programming. Works on both logistic regression and neural networks.
CVXPY
MOSEK
XAI
Explore →
Explainable AI · IEEE CJECE 2026
CERTS — Robust Counterfactuals for Tree Ensembles
MCTS-based counterfactual generation that survives model retraining — ranked #1 in 64% of 248 perturbation scenarios across 5 tree models and 4 datasets.
MCTS
XAI
Robustness
Explore →
Undergraduate Thesis
Indoor Navigation for the Visually Impaired
BLE beacons + ML proximity model + a hybrid Kalman-Median filter for reliable indoor positioning. In collaboration with The Ceylon School for the Deaf and Blind.
Python
BLE
Kalman Filter
IoT
Explore →
Hackathon · 2nd Place SLIoT
Ergomate — IoT-Based Ergonomic Office System
A smart office system monitoring posture, hydration, air quality, and drowsiness via webcam — with user alerts and an HR analytics dashboard.
IoT
Arduino
Python
Dashboard
Explore →