Build high quality ML models quickly using a central Feature Generator Library

At TMNL we experienced some pain when multiple teams were developing features on a huge volume of bank transaction data. We will explain what our issues were and how building a Feature Generator Library helped us to boost ML model development speed and quality.

In our talk we will explain why we built a Feature Generator Libraray at TMNL, what it actually is and how we built it. The focus is on a technical audience but the high level message is useful for POs and strategic data professionals.

We hope to deliver the following takeaways:
- When is a feature generator library useful
- How to build one successfully and what are the tradeoffs to consider