Working as a Teaching Assistant for the course ABE 205 - Engineering Computations For Biological Systems where I help conduct lab sessions involving Data Science and R and grading the assignments.

- Working under Prof. Eugenio Culurciello in e-Lab to develop an agent that can help users by answering questions or performing tasks given by them in indoor environments.- Conducting research on Large Visual Models, including fine-tuning, prompting, and instruction tuning to achieve optimal model performance.- It has various applications such as enabling the robot to remember objects it has seen and inform users of their locations. Additionally, it enhances robot navigation capabilities, enabling the robot to perform tasks such as moving to the kitchen and doing dishes upon user request.

- Worked with Prof. A. Qureshi in the Cognitive Robot Autonomy & Learning Lab (CoRAL Lab) to improve multi-robot motion planning capabilities. My work revolves around the Monte Carlo Tree Search (MCTS) algorithm to efficiently rearrange objects into the desired goal locations while ensuring collision-free navigation for the multiple mobile robots. In addition to this, a strong emphasis is given to incorporating long horizon planning for moving objects within an efficient inference time while also achieving a 2x reduction in the search time.- Developed an efficient motion planner for OMPL that performs time aware collision checking for multi agent tasks.

Working as a Teaching Assistant for AGR 333: Data Science for Agriculture.Developed comprehensive course materials for AGR 33000 - Data Science for Agriculture, including lecture notes, assignments, and practical exercises.Assisted students during computer labs, providing guidance and troubleshooting issues related to R programming and various Data Science problems.Facilitated weekly discussion with faculty members, working collaboratively to continuously improve the course content and effectively manage course activities.

WSO2 is a leading software company in Sri Lanka which is headquartered in Santa Clara, CA, USA.Led the design and development of the Low code suggestion feature that can suggest useful Low code elements that the user might use next, promoting developer productivity by 50%.Experimented with different preprocessing techniques to clean and extract features and researched various Deep Learning models ranging from LSTMs to LLMs such as GPT-3 to provide accurate Low code suggestion and code generation features for the users.Co-developed a fully automated MLOps pipeline for Low code suggestions, that includes preprocessing, model training, evaluation, and serving capabilities driving the periodic ML update cycle from several days to 1 hour.Developed an LLM based “Agent” microservice that can understand natural language queries given by the users and convert them into API calls, even dynamically invoking a sequence of APIs depending on the user's requirement thus improving user experience 10 folds.Designed and developed an event driven application that can provide near real-time alerts for an EIPaas platform enabling a 23% improvement in customers who strive to provide 99% availability SLA guarantees.

Experimented with the automatic tuning of thread pool sizes of Ballerina microservices by taking into account the real-time load configurations and feeding them into the Gaussian Process Bayesian optimization model improving the 99 percentile latencies up to 26%.Conducted research on automatic Database server parameter tuning with the help of existing literary works in that area in the hopes of achieving a potential reduction of cost of operations by at least 10%.