I'm a Data Scientist transitioning from a full-time role in automotive manufacturing, where I spent years optimizing production processes through data analysis.
I completed the Master in Data Science at ProfessionAI while working full-time - every project in this profile was built in parallel with a demanding job, which taught me to work autonomously, prioritize effectively, and deliver results under real constraints.
Before entering the field formally, I already applied data-driven decision making in practice: in my Bed & Breakfast business I used platform analytics to improve KPIs, and in automotive I analyzed process variable correlations to solve production issues.
Languages & Tools
Python R SQL Tableau
Machine Learning & AI
scikit-learn XGBoost TensorFlow Keras NLP Deep Learning Big Data
Libraries
pandas NumPy Matplotlib Seaborn gensim TextBlob PySpark Tweepy
| Project | Description | Tech |
|---|---|---|
| Toxic Comment Classifier | Multi-label toxicity detection with BiLSTM and Focal Loss - AUC 0.97 | TensorFlow · Keras |
| Face Detection | SVM + HOG pipeline with sliding window inference - AUC 0.96 | scikit-learn · OpenCV |
| Creditworthiness Prediction | 291 model configurations compared - Random Forest achieves Recall 1.000 | scikit-learn · XGBoost |
| Spam Detection | Classification + LDA topic modeling + GloVe semantic distance + NER | gensim · spaCy · TextBlob |
| Twitter Sentiment Analysis | Real-time pipeline on Databricks with Spark Streaming | PySpark · Tweepy · Databricks |
| Banking SQL | Feature engineering pipeline - 26 behavioral indicators from relational DB | MySQL |
Open to Data Scientist / Data Analyst positions.