Bristech: MLOps is a one-day event focused on the increasingly important and emerging field of MLOps: the confluence of machine learning models and their iterative, regular deployment into production environments.
We have chosen the term MLOps (over Machine Learning Infrastructure) since it offers a parallel to the (perhaps) familiar concept of DevOps.
Themes explored on the day will include:
The wider context for MLOps: key challenges (such as data provenance) and why it is important
Tooling: including approaches such as Kubeflow, Seldon, etc.
Practitioner experiences: witness & insights from jobbing MLOps practitioners e.g. A/B model testing, multi-arm bandits, etc.
Multiple perspectives: bridging gaps in understanding and terminology between S/W engineers / Ops, Data Scientists / ML Researchers, Managers & Tech Founders.
Maturity of MLOps platforms: how to contextualise the journey towards MLOps nirvana within a maturity framework