I worked as a Business Intelligence Analyst. However, I did more than that title.
CATTHAI is a manufacturer in supporting industry. The company has some subsidiary. I went to a packaging division to manage data flow and support especially in MS Excel. A lot of VBA for automation. And Power Query for data model. But above all, I taught my colleagues basic Excel to boost their proficiency.
When going back to the big factory, it was post-COVID and the AIs hit hard. I and my colleagues started experiencing object detection, with YOLO - at that moment v8 - to support the business. The purpose is to help finding missing parts or error products. Everything starts from zero, with existing resources, I felt the joy on doing r&d like building a toy from scratch. Learned Python, openCV, then the program worked with 0.8 fps, had conversations to ask for better hardware, realized it is not hardware problem, then try again... The most regret part is my contract ended before everything is finished.
And there were a lot of self-learn here. Started from Excel and basic report, I learnt the stack of using Power Query/Power BI for visualization. Then the question is, what if the data is huge? To answer that, I learnt postgresql. Then another question to question that I asked myself or I was asked that push me to learn more:
- The company cannot buy Power Bi license? => switch to Looker, or Apache Superset
- Export data to API? => backend with nodejs
- The other team want to see statistical charts => Python (seaborn) - unfortunately I forgot almost all because I am not really into Python
- The management board want to see charts on website => Quasar Vue to build frontend, API from above then other lib such as D3.js for charts
- The department need to store data onsite and sync => Truenas Scale with 18TB on old server
- And many other things...