[ Tech stack ]
Kubeflow
The ML toolkit that turns a Kubernetes cluster into an AI platform.
Kubeflow brings the ML lifecycle to Kubernetes: Pipelines orchestrates training workflows as containerized DAGs, Notebooks provides isolated Jupyter environments, Katib automates hyperparameter tuning and KServe serves models with autoscaling. Multi-user with per-namespace isolation, it's our choice to industrialize a team's ML on an existing cluster, cloud or on-premise, without vendor lock-in.
[ Why Kubeflow at Dexon ]
What this technology does well, and why we use it.
Typical usage: Shared ML platform on Kubernetes, distributed training, serving.
- 01
Pipelines: containerized, versioned, replayable ML workflows.
- 02
Katib: distributed hyperparameter tuning and NAS on the cluster.
- 03
KServe: model serving with autoscaling, canary and scale-to-zero.
- 04
Multi-user: isolated notebooks, RBAC and quotas per namespace.
[ Complementary technologies ]
The building blocks we often mobilise alongside.
A stack rarely exists alone. Here are the technologies Dexon most often pairs with this one, through pipeline habits, usage similarity or internal mastery. Click on a building block to see its scope.
[ Reassurance ]
- 0+
- custom projects delivered
- 30
- engineers, designers, project managers
- 80 %
- from top French schools
- 24 h
- average reply time
[ They trust us ]
More than 100 French and European companies trust us
[ Press ]
They talk about us.
Nationwide coverage by BFM Business, Le Figaro, Challenges, La Tribune and CNews. An outside reading of our work and our innovations.
