At Neyond, I work on client default and contract risk information for regulated banking. I took end-to-end ownership of data pipelines processing millions of records daily, cutting runtime on key processes by up to 40%, and migrated a replica of the production environment from Hadoop to Databricks.
I also built a personal football analytics app with an expected goals model and a player similarity tool, deployed using Streamlit. In my earlier Neyond role, I built and optimized ETL processes for business banking risk systems and reduced production error frequency from weekly to less than once a month.

