Data & AI Engineer · Trustworthy AI & Governance
Emmanuel Oyelana – building GenAI systems and the data behind them.
Zürich-based. I build LLM systems for banking use cases and governed data pipelines across industries. I evaluate LLM outputs with rubrics mapped to FINMA Guidance 08/2024 and the EU AI Act.
Tools I build with
Python
My first tool for most problems: ETL, LLM integrations, and small services.
Databricks
Lakehouse ETL, Delta tables, and Spark transformations.
SQL
Data modelling, transformations, and quality gates. Still the sharpest tool for all three.
GCP / Azure / AWS
Where my pipelines run: mostly Azure and GCP, with some AWS.
Where I focus
GenAI in Banking
LLMs and NLP for client contact notes, KYC, and fraud detection.
Team winner, RiskON 2025 · Contributor, UZH white paper (2026)
Data Pipelines
Reliable ingestion and transformation in SQL and Python, with Airflow and Databricks where they fit.
Clinical trials at Metronomia · Patient records at GZO · IoT at Nexxiot
LLM Evaluation
Rubric-based scoring of LLM outputs and regression tracking across model versions.
Nine months of production LLM reviews in German and English