Open to opportunities

Sahan Rajapakshe

@sahanrajapakshe

Message

AI/ML engineer building production-ready automation and MLOps solutions.

Sri Lanka

Message

What I'm looking for

I seek roles where I can build production AI systems and MLOps pipelines, focus on NLP and automation, work with containerized, scalable infrastructure, and grow in a collaborative, impact-driven engineering team.

I am an AI/ML engineer focused on delivering production-ready AI automation systems and scalable MLOps pipelines that reduce manual effort and improve model performance.

I have built a SaaS multi-agent business automation platform that cut administrative work by 40–60%, implemented RAG architectures to raise Q&A efficiency by 80%, and developed WebSocket-based real-time conversational systems using Django Channels and Redis.

My model work includes fine-tuning Gemma3 with LoRA for Sinhala and Singlish, achieving 25–30% higher accuracy on grammar tasks and reducing computational overhead by 15%, and developing end-to-end MLOps pipelines for rainfall prediction that improved forecast efficiency by 40% while halving deployment time.

I combine full-stack development skills, containerized deployments (Docker), and experiment tracking (ZenML, MLflow) to create reproducible, CI/CD-ready solutions that scale for businesses.

Experience

Work history, roles, and key accomplishments

AI/ML Engineer

Self Employed

Built production-ready AI automation systems and end-to-end MLOps pipelines that reduced manual work for businesses by 40–60% and improved model accuracy by 25–30%. Delivered containerized, scalable solutions and real-time multi-agent chat systems using RAG architectures.

Python PyTorch Docker Django MLOps RAG WebSockets

MLOps Engineer (Project)

Personal Projects

Developed an end-to-end MLOps pipeline for rainfall prediction that improved forecast efficiency by 40% and reduced deployment time by 50% through ZenML integration and containerized CI/CD workflows.

MLOps ZenML Docker CI CD Python data preprocessing

Machine Learning Engineer (Project)

Personal Projects

Fine-tuned Gemma3 LLM with LoRA to improve Sinhala and Singlish generation by 25–30% and optimized model performance to reduce computational overhead by 15%. Documented reproducible fine-tuning processes for future iterations.

LoRAs PyTorch NLP Model Fine Tuning Evaluation Performance Optimization Gemma 3

Education

Degrees, certifications, and relevant coursework

National Institute of Business Management

Bachelor of Science, Software Engineering

2023 -

Activities and societies: Developed AI/ML projects including a multi-agent automation platform, fine-tuned LLMs for Sinhala, and built an end-to-end MLOps rainfall prediction pipeline.

Pursuing a Bachelor's Degree in Software Engineering with coursework and projects focused on AI/ML, MLOps, and full-stack development; expected graduation in September 2027.