Anton Price

This role was a 1 year fixed term contract (extended by 1 quarter) as a Data Scientist in order to provide cover for a colleague who is on maternity leave. The responsibility has been primarily around managing the ETL and financial reporting, most notably around invoices. The work primarily revolved around using Python and SQL to manage the data engineering and maintaining/running those jobs. I have also developed some new automation in order to deal with the high volume of requests which were previously done manually.

Upon first joining AnyVan there was no data science at all in the business and I was tasked with developing a ML model for customer pricing. This was a challenging role as it involved a lot of process design and development from scratch as well as setting up development environments (in AWS cloud mostly) and assisting engineering in the integration with the existing business processes. I came up with a novel 2-stage pipeline involving ensembling a GBM with a transformer neural network, which was successfully deployed for live pricing. I also used snowflake/DBT to do the data extraction and some feature engineering.

Within DLG I worked in the Darwin brand as a part of the pricing team and was primarily responsible for driving improvements to the motor pricing model stack (e.g. via feature engineering/improved feature selection) and developing a fraud propensity model to enhance the business's counter fraud capabilities. As there was a very small P&U team, I was also responsible for day to day pricing queries and investigating discrepancies from customer facing colleagues, tuning the ML models and supporting governance procedures around model documentation and sign off.