Chamin Hewage (Ph.D.)

Senior Software Engineer (AI & ML)

✉ chmk90@gmail.com 📱 +353-89-472-6069 🔗 LinkedIn 📝 Medium 🤖 HuggingFace

Skills

Languages:
Python, Java, SQL, JavaScript, Node.js, HTML/CSS, PHP, Shell scripting
AI/ML:
Neural Networks, LLMs, Context engineering, RAG, Agents, Fine-tuning, AI Evals, AI/ML model testing, n8n
Technologies:
Git, Kubernetes, Pandas, NumPy, Scikit-Learn, Spark, PyTorch, Pydantic, Transformers
Domains:
Distributed Systems, Databases (SQL & NoSQL), System Design, DataLakes

Professional Experience

Hewlett Packard Enterprise (HPE)
Senior Software Engineer (AI & ML)
Ireland
Aug 2023 - Present
  • Build data integration pipelines to ETL from diverse sources, achieving 3–4x scalability, 55–80% performance gains, 60% faster deployments, and $250K+ savings across 6+ major analytics services.
  • Designed and developed an AI-native IP licensing platform with secure vault storage, time-bound encryption, and monitoring to align with company strategy and enable innovative IP monetization. (on-going)
  • Implemented inference optimizations (async calls), evaluation techniques (BLEU, ROGUE, Human evals), LLM-based layout extraction and OCR-based text extraction, and LangChain to innovate legal contracts chat bot
  • Implemented multi-agent, multi-model (text and vision) based AI search engine prototype to reduce time and human capital allocated on RFPs at prominent innovation hackathon at HPE
  • Interviewed (contract) candidates, established engineering practices, mentor junior engineers, review HPE white papers, and deliver company wide tech talks on cloud architecture and GenAI
  • Co-authored IP disclosure for HPE's Cloud Platform portfolio (under review)
University College Dublin
Scalable data systems researcher [PhD]
Ireland
May 2019 - Apr 2023
  • Architected a scalable cloud data lake to integrate large data sets in Hadoop and HBase environment
  • Innovated scalable three index structures, for distributed databases, to achieve a 34-55% improvement in storage, a 3.4-4.7x gain in query performance, and a 3-5.2% enhancement in CPU execution
  • Contributed to implement a Spark-based distributed batch-processing algorithm to speed-up batch processing 16x
  • Presented research globally (New York, San Fransisco, Sydney, Osaka, Amsterdam, Paris) and won best research presentation at an IEEE workshop
Cisco Systems Inc. & Pi School
Machine learning summer intern
Italy
June 2018 - Aug 2018
  • Built an ML pipeline to enable application of different ML models which implemented through Scikit-Learn
NTU
Software Engineer
Singapore
Feb 2017 - Apr 2019
  • Implemented Markov Model based real-time data monitoring app with pattern recognition and anomaly detection
  • Led migration of SQL-based full-stack application from monolithic to microservices while implementing REST APIs
Sysco LABS
Software Engineer
Sri Lanka
Feb 2016 - Feb 2017
  • Developed remote-debug feature to reduce the debug effort time from ~10 minutes to 30 seconds
  • Contributed to developing many core features for the full-stack web app that server 4000+ unique users

Generative AI / Reinforcement Learning Projects

ICØN Finalist - AI Ireland, 2025

AI Identity management platform (Claude, 11Labs)

DeepSeek demo
TechIreland AI challenge demo
AI content search

Multi-agent, multi-model top k image recommendation with guardrails (LangChain, BLIP vision models, FLAN-T5-XL model, dynamic model selections, semantic understanding, etc.)

Flight Simulator

A browser game based on Q-Learning and Basic Agent reinforcement learning.

FluxDev-Fine-tune

Applied LoRA fine-tuning and quantization atop stable diffusion based Flux-dev model to generate custom images while ensuring affordable fine-tuning on lower VRAM.

BrailMenuGen

Generate Braille-friendly PDFs from images (OpenCV, LlaMA-3-8B, LayoutLMv2)

Education

PhD in Computer Science (Scalable integrated data systems)
University College Dublin
Dublin, Ireland
May 2019 - Apr 2023