Chathura Rathnayake

I’m a hardworking, focused AI enthusiast with strong machine learning theoretical understanding and industrial exposure. I enjoy solving real problems with analytical thinking, and I’m motivated to pursue a Machine Learning career while contributing to my organization and humankind.

At ISA (Air Arabia Group), I’ve played a key role designing, implementing, and deploying FastAPI microservices hosted on K8 clusters, including static and dynamic input feature services. I’ve built data monitoring and data-quality dashboards, developed build curve computation logic, and created a workflow orchestrator for end-to-end ML pipelines. I’ve also led distributed model training and inference using Ray, migrated production infrastructure to an Airflow-based system, delivered passenger demand forecasting end-to-end, and built flight anomaly detection using a modified Z-score approach—currently extending it into a multi-agent anomaly investigation system for root causes, business explanations, and recommended actions with production guardrails.

Before that, at Yaala Labs, I engineered an LLM-based chatbot on AWS using Retrieval Augmented Generation to answer internal knowledgebase queries, and I supported day-to-day product tasks. I also built an AI code reviewer, contributed bug fixes to the LangChain LLM framework repository, implemented database tampering detection with cryptographic controls across DynamoDB, MongoDB, and PostgreSQL, and supported mission-critical components through contributions like the P8 Digital Assets Exchange. Earlier, I completed a research internship at the University of Colombo exploring Multisensory Augmented Reality using Google ARCore.