I’m Bima — a former banker who swapped spreadsheets for Python scripts, and now I’m deep into the world of data analytics (with less paperwork and way more coffee ☕).
After leading high-performing teams in the banking world, I decided it was time to take my love for numbers to the next level. I jumped into a 12-week data science bootcamp where I wrestled with messy datasets, built machine learning models, and made dashboards that even my non-tech friends could understand.
Collaborated on the development of Wangi, an AI-powered perfume recommendation app that uses NLP and TensorFlow to match user scent preferences with suitable products based on real customer reviews. Designed to enhance fragrance discovery for consumers and provide actionable insights for beauty brands in the Indonesian market.
- Tech stack: Python, Pandas, Matplotlib, Seabron, Numpy, Selenium, Beautifulsoup, NLTK, Googletrans, Scikit-learn, Sentence Transformers, HuggingFace
⛅ Weather Image Classification Using EfficientNetB0 (CNN)
Developed a CNN-based model for weather image classification to support early warning systems in sectors like transportation and agriculture. Designed for deployment in remote areas with limited infrastructure, the model enables real-time, automated 24/7 weather reporting as a complementary solution to existing forecasts.
- Tech stack: Python, Tensorflow, Keras, Matplotlib, Seaborn, Scikit-learn, Pandas, Streamlit, HuggingFace
📈 Market Report on Revenue Driving Factors in the Restaurant Industry
Conducted end-to-end analysis to identify key factors influencing restaurant revenue. Simulated real world workflows by ingesting Kaggle-based data into Elasticsearch via automated pipelines built with Apache Airflow, combining Data Analyst and Data Engineer responsibilities to generate business insights and strategic recommendations.
- Tech stack: Python, Pandas, Numpy, Scikit-learn, Apache Airflow, PostgreSQL, ElasticSearch, Kibana, Great Expectations
☀️ Mental Health Analysis: Predicting Depression in Students with Machine Learning
Developed a machine learning model to predict students at risk of depression using survey-based data, enabling early identification and supporting targeted mental health interventions.
- Tech stack: Python, Pandas, Numpy, Scikit-learn, Matplotlib, KNN, SVM, Decision Tree, Random Forest, CatBoost, Streamlit, HuggingFace
🏬 Supplement Sales on E-Commerce Platforms
Conducted exploratory data analysis and visualization of supplement sales across major e-commerce platforms (Amazon, Walmart, iHerb) using Pandas, uncovering sales trends, pricing insights, and brand performance through statistical techniques.
- Tech stack: Python, Pandas, Scipy, Matplotlib, Tableau
Whether it’s crunching numbers or swapping insights, my inbox is always open — let’s talk data!