Building practical AI and data systems with Python, machine learning and embedded technologies.
I'm an AI Engineer focused on Python, Machine Learning, Data Engineering and AI applications.
My experience includes building data pipelines, machine learning models, LLM-based applications and embedded systems connected to cloud services.
I've worked with technologies such as Python, PyTorch, TensorFlow, PySpark, AWS and C++, across projects involving data processing, AI, geospatial analysis and Edge AI.
I enjoy working across different parts of a system, from data acquisition and processing to AI models and deployment.
- Machine Learning & Deep Learning
- Generative AI & LLM applications
- Retrieval-Augmented Generation (RAG)
- Data Engineering & ETL
- Python backend development
- Geospatial and satellite data
- Edge AI & Embedded Systems
- Cloud-based AI applications
Geospatial application for wildfire risk assessment using Sentinel-2 satellite imagery.
The project retrieves and processes multispectral satellite data, applies image preprocessing and calculates spectral indicators to support fire-risk analysis.
Highlights
- Sentinel-2 satellite data
- Geospatial processing
- Remote sensing
- Spectral indices
- Automated data acquisition
- Environmental risk analysis
Tech
Python Rasterio NumPy Copernicus STAC Remote Sensing
Machine learning platform for analysing electricity consumption, renewable energy and CO2 emissions.
The project combines data analysis, feature engineering, machine learning and deep learning models to study energy consumption patterns and forecasts.
Highlights
- Data preprocessing and analysis
- Feature engineering
- Machine learning models
- Deep learning
- Time-series analysis
- Data visualization
Tech
Python TensorFlow Scikit-learn Pandas NumPy
Experimental prototype exploring LLM interaction, RAG and tool-based actions through a 3D interface.
The project connects a Python backend with a Three.js environment, allowing an AI assistant to interact with elements of a 3D scene.
Explored
- LLM-based interaction
- RAG
- Tool calling
- AI orchestration
- 3D interfaces
- Local and cloud LLMs
Tech
Python Flask LLMs RAG Three.js
Python-based document generation pipeline for creating consistent, ATS-friendly resumes from structured data.
The system uses a deterministic workflow to validate resume data, generate HTML and produce the final PDF.
Highlights
- Structured YAML input
- Data validation
- Automated document generation
- HTML templates
- PDF rendering
- Reproducible output
Tech
Python Jinja2 Playwright YAML
Python PyTorch TensorFlow Keras Scikit-learn Hugging Face
LLMs Generative AI RAG Prompt Engineering
PySpark Pandas NumPy SQL PostgreSQL MongoDB
ETL Web Scraping
FastAPI Flask REST APIs PyTest
Docker Git Linux
C++ ESP32 FreeRTOS MQTT IoT
AWS
Worked on private AI and data-related projects involving Python, Machine Learning, Deep Learning and data processing.
Developed and integrated AI components using Python and frameworks such as PyTorch and TensorFlow, while working with PostgreSQL and Docker.
Worked on the embedded side of an AI wearable prototype.
- Developed C++ firmware for ESP32-S3 using concurrent tasks.
- Integrated microphones and sensors for contextual data acquisition.
- Implemented MQTT and REST communication between embedded devices and cloud services.
- Contributed to the integration of AI capabilities within the device architecture.
Worked on automated data collection, processing and ETL workflows.
- Developed multithreaded web scraping systems using Python and Selenium.
- Built ETL pipelines with Python and PySpark.
- Processed large volumes of geolocated data.
- Worked with MongoDB and AWS services.
- Implemented automated testing with PyTest.
Co-founded Crazy Devs 8 and developed Unity projects using C#.
Worked on gameplay systems, enemy AI, pathfinding and rapid prototyping in small development teams.
Game development is no longer my main focus, but it provided a strong foundation in software development, AI behaviours and collaborative engineering.
