- Develop the posting system and matching system of buyer demand for real estate agent, including frontend to backend, and most of feedback. It has more than 500 demands on our system in first two months. Use government's API to replace Google's geocoding API and help us to save 20,000 NTD a month - Responsible for moving 18,091,403 documents of meta data from Datastore to MongoDB and speed up our data pipeline about 3 to 4 times. Improve CRM for internal use, design new data architecture for organizing data more correctly, and speed up the performance. - Cooperate with colleague to help upgrading python from 2 to 3, speed up most of our repository. - Being many different roles in each project, helping bug fix, and knowledge sharing.