Chalitha Perera Email & Phone Number

Sri Lanka

Colombo, Western Province, Sri Lanka

In my current role at CUBE Global, I collaborate with a team of experienced data scientists to develop NLP services that extract key information from regulatory documents, effectively turning complex data into actionable intelligence. I have been responsible for designing NLP service development and deployment architectures and aiding in the transition of machine-learning models from proof-of-concept to production services.- Led the Migration to Kubernetes: Directed the shift of NLP services and machine learning models to a Kubernetes-managed environment, using FastAPI for standardized micro-service templates. This established a highly scalable, reliable production-grade AI infrastructure.- Implemented MLOps and LLMOps Practices: Spearheaded the adoption of MLOps and LLMOps practices across AI teams, automating and optimizing the pipelines from development to production. This initiative significantly reduced manual processes and accelerated deployment times, boosting overall team productivity.- Improved Data Annotation Processes: Improved the data annotation process by transitioning to a Label Studio setup from a previously inefficient Excel-based system. Developed custom annotation UIs for Label Studio, which enhanced productivity and improved the quality of annotations.- Optimized AI Services: Led the optimization of several AI services and models for production, significantly enhancing runtime performances.

At Rozie AI, a startup specializing in Conversational AI for automated customer experience, I served as the main NLP engineer heavily contributing to the design and development of the Rozie AI’s Natural Language Understanding (NLU) engine. My responsibilities encompassed training and deploying machine learning models and services for intent classification, concept tagging, sentiment and emotion extraction, semantic search, and question answering. Throughout my tenure at Rozie AI, I played a critical role in developing and optimizing AI-driven solutions that enhanced customer service interactions, leveraging natural language understanding to automate and personalize customer engagement. This position allowed me to refine my skills in designing and deploying robust AI solutions in a challenging and dynamic startup environment.

Colombo, Sri Lanka

In my role at Zaizi, I was involved in developing Sensefy, an enterprise search engine with semantic and entity-driven search capabilities. My contributions were focused on developing machine learning models for document classification, clustering, and topic modeling, which delivered significant value to Alfresco clients. During my tenure, I also collaborated with researchers from leading European universities and research institutes on the Mico Project, a cross-media analysis research initiative. This experience not only enhanced my technical skills but also allowed me to contribute to high-impact research and development projects.

Worked on integration of YAGO with Stanbol and implementation of named entity disambiguation engine for Stanbol

Colombo, Sri Lanka

Worked on applying machine learning techniques for automating enterprise content management. set of tools are implemented for document clustering, classification and topic modeling using Apache Mahout

B.Sc. Eng (Hons), Computer Science And Engineering, Second Class, Upper Division

G.C.E. Advanced Level, Physical Sciences (Mathematics, Physics, Chemistry)