- Navigation Service - GPS raw data analysis - Developing speed aggregation and prediction pipeline - Creating service metrics and service monitoring - Developing user & road level speed aggregation pipeleine - Modeling for future road speed forecasting- Demand / Supply modeling - Forecasting features with statistical methods and time series methods in variety of services- LLM Service - prompt engineering and developing serving architecture - RAG, search optimization chatbot - Fine tune LLM for specific tasks- Safety driving score development

Factory Anomaly Detection (20.07 - 20.08)- Unsupervised learning and convert excel data into time series dataset- Used PyTorch GRU Auto Encoder and time series clusteringWashing Machine Vibration Optimization (20.09 - 21.04)- Used Xgboost, Bayesian Optimizing to make prediction score table- Used python fastapi to construct full api framework- Validation by actual experiment in laboratoryCustomer Pattern Prediction (21.05 - 22.01)- In Hadoop database, constructed data mart for analysis- Used LightGBM, Bayesian Optimizing to predict customer pattern- Used Logistic Regression to make scorecard for scoring individual customers- Ensemble Distilbert, TabNet to predict customer patterns and prioritize variables

• Database setup by collecting the datasets from the startups and made it as one piece by using SQL• Data cleaned by using R and visualized it in Tableau to make it as an interactive plot that is more understandable

•Teaching students English and Math with English(usually go over every materials that students have done in their school.)

•Data-mining by using DART, used that data to suggest business strategy for sales promotion•Designed project proposal of CRM data from company’s application by comparing all the applications available on apple store and report what can make our application more user-friendly and informative.•Learned how to read financial statements and eject useful criteria to measure company’s status.