AWS re:Invent 2019: Building machine-learning infrastructure on Amazon EKS with Kubeflow
During re:Invent 2019 I joined my colleague Jean-Marie from Babylon and Yaniv from AWS on stage to discuss MLOps, using AWS Kubernetes service: EKS. The talk covers some of the challenges of managing ML Infrastructure in an organization, and how technologies like Kubeflow and Kubernetes can help and do the heavy-lifting.
Speaking at AWS re:Invent was a great experience, and I am thankful for the opportunity that was given to me. I have to say that I was surprised with the event overall. I was expecting a lot of vendor-sponsored talks, but it turns out most of the breakout sessions I watched were deeply technical, of exceptional quality, and didn’t feel like someone was trying to sell you something. People were on stage, discussing the engineering problems they experienced, and explained how they solved them. I learned a lot that week. Would definitely recommend.
KubeCon NA 2019: Building a Medical AI With Kubernetes and Kubeflow
In late 2019, I was selected to present at KubeCon CloudNativeCon North America in San Diego, the biggest Kubernetes conference in the world, and talk about the work I’ve been doing at Babylon Health: building a Machine Learning platform on Kubernetes.
Minimum Viable Docker
This was my first talk as an engineer, back in 2017, when Babylon Health was running containerized services without tools like Kubernetes. That system worked very well for us, until Babylon started scaling really, really fast. When that happened, we moved to Kubernetes.
It is now 2019 and we are running hundreds of services on top of Kubernetes in Europe, Asia and America.
While Kubernetes and Mesos are all the rage, you don’t necessarily need a complex orchestration layer to start using and benefiting from Docker. We will present how Babylon Health is running its dockerised AI microservices in production, pros and cons, and what we have in store for the future.