- Conduct weekly office hours to assist students with course concepts, answer questions, and provide debugging support.- Proctor exams and grade assignments for a cohort of 150 students.

- Developed a data pipeline to preprocess and analyze a dataset comprising fire hazard reports and community fire history records. Constructed a model pipeline by stacking several ensemble models to predict the next 7-day fire risk.- Employed Shapley Values to conduct a causal analysis, identifying critical hazard characteristics influencing fire rates.-Built RESTful API endpoints using Django to send daily predictions and causal analysis, facilitating real-time access to valuable insights for informed decision-making.- Developed an interactive web map using React and Leaflet to visualize community fire predictions dynamically.-Deployed the pipeline with an 80% recall and 71% accuracy, resulting in the project being implemented operationally in 111 communities in Longgang District, Shenzhen. A patent was also obtained for the solution. Additionally, secured a government grant of 100,000 CNY for the project.