In Episode 13 of Season 3, of Driven By Data: The Podcast, Kyle Winterbottom is joined by Leanne Fitzpatrick, Director of Data Science at Financial Times, and Andy McMahon, Head of MLOps at Natwest Group, where they discuss the current MLOps maturity, which includes;
- The maturity of MLOps over the last 12 months
- Why getting ML into production is no longer the challenge
- Why the Ops component is now the major challenge
- What best practice actually looks like
- Why ML is more software than Data Science
- Why thinking about MLOps as products is helping the industry
- What does MLOps actually mean to business leaders
- The resiliency and risk elements
- Why design and processes are really what make or break MLOps
- Why we need rules of engagement
- Why we don’t have enough practitioner expertise
- The challenge of building scalable MLOps solutions
- The differences and similarities to software engineering
- Why companies will need to change their operating models to support their MLOps maturity
- The importance of domain expertise for successful MLOps
- The current tooling landscape
- The role of vendors play in the journey
- Build versus Buy
- Predictions for next 12-24 months
Driven by Data: The Podcast