Benjamin Calvet
Cern, Computing Engineer for CMS Experiment.
Genève, Switzerland
Actions
I am Benjamin Calvet, originally from the south of France, deeply passionate about the CNCF ecosystem, with a special focus on Kubernetes. I became a Kubestronaut last year, reflecting my dedication and expertise in this field. I am particularly interested in technological innovations and optimizing the tools we use, especially in a context where resource rationalization and consumption reduction are key challenges.
Area of Expertise
Topics
Kubernetes is Awesome… But Do You Actually Need It?
Kubernetes is the de facto standard for container in the orchestration industry. It's scalable, resilient, and packed with fantastic features and a massive ecosystem.
But what if your workload is a single microservice ? A hobby project? Or a team of three people ?
In this talk, we’ll explore the hidden costs of Kubernetes : operational complexity (especially on-prem), steep learning curves, and the "just use K8s" cargo cult.
We will demonstrate with real world use case how Kubernetes can be sometimes a little bit too much for the use case.
Here's are the topics that we'll will dive in :
- Use case breakdowns: From small-scale apps and edge computing to prototyping, where lighter alternatives like Nomad, ECS, or even bare-metal solutions deliver better results.
- The Kubernetes overhead : Maintenance overhead, data persistency, security overhead.
- How to push back when stakeholders demand K8s for the wrong reasons, propose a decision "framework".
- The alternatives we have (e.g., Fly.io, Railway, CapRover, or even no orchestration at all).
Attendees will leave with actionable strategies to "right-size" their tooling, avoiding the "Kubernetes everything" trap. Whether you're a dev, ops engineer, or architect, this talk promotes smarter, more efficient cloud-native practices—because sometimes, less is more.
From Kubernetes to Photons : Orchestrating Multi-Agent AI for Entertainment
Multi-agent architectures usher in a new era: instead of a single monolithic AI model, we deploy a set of specialized agents that collaborate, each mastering a specific task, to form a globally more robust and scalable system.
But how can we leverage Kubernetes orchestration to create the future of distributed AI ? Inter-agent communication (sync, event, async), new orchestration patterns, and service exposure (RAG, MCP) are the new challenges.
This session explores the similarities and differences between cloud-native orchestration patterns and those required for deploying AI agents.
As crazy developers who love great music and entertainment, we created a case study to demonstrate how distributed AI on Kubernetes can create immersive ambiances by controlling sound and lighting effects.
This hands-on demonstration will show how to reconcile Kubernetes and distributed AI back into the physical world, bridging cloud-native orchestration with immersive hardware-driven experiences
Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.
Jump to top