Independent Data & AI Engineer focused on data engineering, advanced data cleaning, applied machine learning, and the development of AI systems.
I build reproducible data pipelines, high-quality datasets, predictive models, AI agents, and language models from scratch, with a strong emphasis on transparency, experimentation, and practical implementation.
My work spans:
- data integration, cleaning, validation, and quality control;
- exploratory analysis and reproducible research;
- feature engineering and predictive machine learning;
- explainable AI and model interpretation;
- AI agents with routing, planning, tools, memory, and human escalation;
- agentic AutoML and experiment-driven ML systems;
- language models and Transformers implemented from first principles;
- analysis of public, socioeconomic, institutional, and climate data.
I primarily use Python to transform complex data into reliable, understandable, and actionable systems and insights.
Expert in Datasets and Notebooks
- Datasets: Top 30
- Notebooks: Top 511
- Multiple Kaggle medals across datasets and notebooks
| Project | Area | Highlights |
|---|---|---|
| Building Agentic AutoML | Agentic AI / AutoML | Experiment-driven AutoML system evolving from a simple baseline to a senior ML agent |
| Customer Support Agent | AI Agent / Human-in-the-Loop | Customer-support agent with classification, automation, and human escalation |
| Building AI Agent | AI Agent from Scratch | Routing, planning, parsing, tool execution, memory, and agent skills |
| Building LLM | LLM from Scratch | Progressive implementation from statistical language modeling to a decoder-only Transformer |
| Global Gender Gap in Education - 1950-2015 | Global Data Analysis | Population-weighted analysis across 146 countries, age groups, regions, attainment levels, and income groups |
| Global Emissions & Temperature - 1950-2024 | Climate / Time Series | 75 years of CO₂, greenhouse-gas, and global temperature analysis |
| Home Credit Default Risk | Machine Learning / Credit Risk | Feature engineering, XGBoost, Optuna, and SHAP interpretability |
| Used Car Prices | Regression / Machine Learning | Feature engineering and LightGBM for used-car price prediction |
| House Prices | Regression / Machine Learning | Ridge regression, skewness transformation, and advanced feature engineering |
- Data engineering and data quality
- Advanced data cleaning
- Applied machine learning
- Explainable AI
- Agentic AI and AI agents
- AutoML and experiment automation
- Natural Language Processing
- Language models and Transformers
- AI systems from scratch
- Public and socioeconomic data
- Climate and time-series analysis
- Reproducible research
