Making sense of ML Observability

Tonight we'll be hearing from:

• Antoni Shtipliyski on "Making sense of ML Observability"
• ML is an invaluable way for Skyscanner to serve the world’s travellers. From the early wins till now they've needed to deploy ever more sophisticated ML models to a greater number of areas. With this came the realisation of the importance of operational rigour. This was what started them on the path to understand why ML observability is important and how to do it effectively.
• This talk walks you from the early days to the present in the hope that it can help other companies in their journeys of making sense of what good ML observability looks like in their own context and what pitfalls they can try to avoid.
• Antoni is an Engineering Manager at Skyscanner where he leads the teams responsible for Machine Learning Operations. Previously he worked in operations and R&D for the upgrades to the Level-1 Trigger of the CMS experiment at CERN which he did as part of his PhD studies with Imperial College London.

We will be graciously hosted this evening by the Bayes Centre for the first time. Many thanks to them!

Plan for the evening:

• 18:30: doors open
• 18:30 -> 19:15: chat / pizza
• 19:15 -> 20:00: talk + q/a
• 20:00 -> 20:30: tidy-up / head to nearby bar
• 20:30: event close