Building practical data pipelines, process mining models, and automated AI tools.
flowchart LR
A["📦 Data Sources<br/>(SAP ERP, Web Feeds, Media)"] --> B["⚙️ Pipelines & Analytics<br/>(Celonis EMS, Databricks, dbt)"]
B --> C["🤖 Applied AI<br/>(Delay ML, Local Whisper, LLMs)"]
C --> D["🎯 Outcomes<br/>(Reduced Delays, Studio Audiobooks)"]
Enterprise Order-to-Cash (O2C) delay prediction and process intelligence engine. Built with Databricks, predictive machine learning on SAP data, and semantic contract retrieval to highlight delivery bottlenecks.
Databricks • Celonis / SAP • Python • XGBoost • RAG
Local multi-agent pipeline that downloads YouTube videos or playlists, transcribes speech with GPU-accelerated Whisper, runs an LLM editorial loop, and generates mastered audiobooks.
Whisper large-v3 • AMD Vulkan • Ollama • FFmpeg
Modern data engineering pipeline modeling vehicle market data into dimensional star schemas using dbt and DuckDB.
dbt • DuckDB • SQL • Python
Curated collection of standalone Python scripts, including a real-time 3D spinning donut terminal animation, an image-to-ASCII converter, and CLI utilities.
Python • Algorithms • CLI Tools
- Process Mining: Celonis EMS, PQL (Process Query Language), Action Engine, Event Log Modeling
- Data Engineering: Databricks, Apache Spark, dbt, DuckDB, Snowflake, PostgreSQL, Docker
- Languages & Tools: Python (Pandas, NumPy), SQL, Git, Linux / Bash