It was a big challenge to work on a Multimodal Sensing Kitchenware Project with Bosch over my summer. I independently developed a time series prediction model using denoising and clustering techniques. After numerous explorations and self learning, I managed to achieve a good performance of less than 30 seconds difference away the desirable grade. With this new model, I scaled down the burden of processing unit with a smaller model size around 2kB that helps to reduce the operating cost by 50%. This sets the foundation for the automation of the machine to stop when reaching the desirable state.

Improve Whatsapp Chatbot features that streamline communication between patients and clinics.Designed conversation tracking dashboard with manual interventionImplemented security protocols in the backend database such as Auth to enhance reliability of storing patient data.

Developed a user friendly website with Reactjs, MongoDB, Express.jsIntegrated Chatgpt API to generate relevant and comprehensive department goals from organisation goals

- Built a portion of the Recommender System that automates the customer recontract process and maximizes the chances of existing clients accepting the recommended offers. Analyzed multiple data structure to store and edit the huge pool of rules in the Recommender System in the most dynamic way.- Ideated a user interface which helps the marketing team to more efficiently input marketing rules into the Recommender System that handles more than 100 combinations of criteria. Created multiple versions of prototypes in Excel and Streamlit for proof of concept.- Customized an anomaly detection model independently using SARIMA Forecasting to help Product Team analyze daily network data and notifies the team when action and attention is needed. This solves Product Team's business challenge having to constantly monitor 200 customers' network usage traffic daily previously.

Extermly fortunate to be able to intern under the Space System Project Division. I led a research project by myself on image enhancement and computer vision using python programming language. Furthermore, I also got a chance to work on a side task relating to machine learning and object detection using MASK-RCNN and Tensorflow.

Conduct workshops and leadership camps for secondary school students.