Michael Levan👨🏻‍💻☕️

An SRE Agent needs 4 things: 1. A proper prompt/system message. This will give the direction for the Agent to take. 2. Agent Skills that are designed to turn your Agent into an SRE and observability expert. 3. A good LLM 4. Targets Number 4 is where the "magic happens". When you put together a solid Agent for a particular use case, the target can be anything in its speciality. For example, in an SRE Agent, the target could be anything from a cloud provider to a Kubernetes cluster. #agenticai #ai

Remember Jev?

Nifty. Agent Substrate calculator for seeing how quickly sessions will resume under specific levels of load. #agenticai #ai #kubernetes

Diving into things at Agent Substrate day! Thanks so much to the organizers for having me on stage to tlak about Substrates components and why they're important for the agentic era. @CloudNativeFdn @KubeCon_ #agenticai #ai

So are NVIDIA GPUs going to be cheaper now or?

This used to bug me big time (@claudeai adding a bunch of comments). Clean code best practices are "of course add comments, but not the comments you don't need". The reason why we never added a crap ton of comments in a codebase before is because it would drastically slow down humans because they had to read through it all. It's not like comments slow the runtime/compilation down though, so... In the world of agentic programming/coding, I'm beginning to think Claude does it so when Claude look at the codebase a week from now, a month from now, a year from now, it has reference points. #agenticai #ai #kubernetes

Sunday funday on the plane heading to @CloudNativeFdn @KubeCon_ A new, and very-much necessary implementation for Agent Substrate installations 🤓 Who's at KubeCon? Hit me up! #agenticai #ai

4 talks over 4 days 🎉 First stop, SREday ! I’ll be diving into all things architecture, isolation, and governance with agent-substrate Next stop, @CloudNativeFdn #KubeCon

Labs have been updated for Agent Substrate v1.0.0 #agenticai #ai #kubernetes

Agent Substrate v1.0.0 is officially out. There's A LOT to a GA release of any software, so I wanted to map out the 5 top-of-mind things. 1. Rolling upgrades are supported 2. One actor's memory use can no longer take down its worker 3. Suspend and resume recover from more failures 4. Egress and router fixes 5. Easier to get started and connect clients #agenticai #ai #kubernetes

v1.37 on AKS == Agent Substrate time 😁

All Things AI Today xitter.twitterapi.xyz/i/broadcasts/2OdXErrpn…

The Agentic Field guide currently has 16 hands-on labs for sandboxing and isolation. Two sections: ✅ Isolated Environments With Agent Substrate ✅ Isolated Agents With Kagent + Agent Substrate This field guide will be updated weekly, so if you're looking to learn more about agentic sandboxing, isolation, governance, and how it works with Agent Substrate, this is the place for you. agenticfieldguide.ai/ #agenticai #ai #kubernetes

100%. Budgets, cost optimization, and known spending makes sense for just about every enterprise. For those who don’t work at frontier labs, tokens get pricy.

In any production implementation, there are two parts of the stack: 1. The underlying platform/system (like the engine) 2. How you're interacting with the system and the protocols that are used One convo I have all of the time is A2A (despite a lot of talk around not knowing the future of A2A) as this is the primary communication protocol between Agents. Agent Substrate may not have the protocol (and for good reason - it's the engine/lower-level primitive for isolation and sandboxing), which is why using kagent to implement protocols like A2A (and various other things like policy enforcement, end-to-end traces, etc.) is needed for the entire agentic stack to come together. With 1 and 2 working together, you only have bits and pieces of what ultimately needs to be implemented in production. #agenticai #ai #kubernetes

I spoke at the AAIF Boston meetup last night hosted at Microsoft R&D in Cambridge. I’m always reminded when I go what an amazing place Cambridge is. You’re surrounded by tech. MIT right down the street. Google HQ in the center. Great vibes. #agenticai #ai #kubernetes

There’s one thing that AI should absolutely be used for. To create, at scale, deterministic implementations. Code generation can still feel janky, but you know what AI is really good at? Following a specific direction. Example: “Use AI to create rules for X system as the need for the rules populate”. Instead of: “Go create me X application. Good luck”. Use AI to create the deterministic tool, not for constant inference.

Agent Substrate, by design, is a lower-level system. It's not supposed to have things like: - Native A2A support (A2A handlers, Agent Card discovery, or A2A message/task APIs, etc.). - Durable Agent history (sessions/memory/context/conversation) - Agent runtimes (vs you having to BYO them) - Runtime auth - The ability to see end-to-end Agent traces and other observability data And that's totally okay because of what it's designed to do. @kagent_dev is designed to implement all of these things for you. Combining kagent + Agent Substrate, you get the isolation/sandbox/security/scale and the Agent support/permanent storage/auth N/Z/observability for your agentic workloads. #agenticai #ai #kubernetes