Join us on Tuesday, 6/23 for the Plural x Kubernetes June meetup 👋 ​Our guest speaker is Adna Zujo Lakisic. Her topic is "Accelerating Multi-agent Development on k8s with Kagent and Mirrord" ✅ RSVP at: luma.com/r5tvqerq 💡 As organizations move from single-agent applications to multi-agent systems, development becomes increasingly difficult. A single workflow may involve multiple agents, tools, services, and APIs distributed across Kubernetes environments. Debugging these interactions often requires repeated deployments and lengthy feedback cycles. Using kagent and mirrord, we demonstrate how developers can run agents locally while connecting to live Kubernetes services, enabling rapid iteration, debugging, and validation of distributed agent workflows without redeploying every change. 🗓️ ​Schedule: 6:00pm - door opens 6:30pm - intros (please arrive by this time!) 6:40pm - speaker programming 7:20pm - networking 8:00pm - event ends See you next week!

Can't wait for #KCDNewYork this Wednesday, June 10th! #Kubernetes #CloudNative

Is it already June?! We wrapped up May with a doubleheader that drew lots of engagement from the audience. Takeaways from @brunochauvet's talk 👇 1️⃣ Fragmentation is the real on-call cost. The data exists, it is just scattered across k8s, observability, CICD, etc. 2️⃣ Combining MCP tools behind a single Gateway tames complex environments. One protocol, one auth boundary, one tool surface. 3️⃣ Autonomous ReAct turns Root Cause Analysis into self-service. Observe → Reason → Act → Verify → Output. 4️⃣ Auto-remediation needs guardrails and a high level of confidence. Focus on known failure classes (OOM, CrashLoop, ImagePull) and expand from there. 5️⃣ A self-improvement loop is what compounds. Agents should review their own investigation logs, propose prompt edits and regression-gate the changes against real production alerts. 🤖 We also watched a demo from @michaeljguarino on workbenches, Plural's always-on agentic AI solution for a variety of DevOps use cases, including incident root cause analysis, cost management, Kubernetes upgrades, self-service, and more. P.S. For those going to @KcdNewyork - drop us a note because the Plural team will be there!

Last call for the Plural x Kubernetes May meetup this Thursday, 5/28! If you haven't signed up yet, please RSVP ASAP to save your seat: luma.com/g0ywmvfq ​Our guest speaker is Bruno Chauvet, Cloud Infrastructure Tech Lead at Rokt. His topic is "No More War Rooms: Kubernetes MCP Servers Agentic AI for Autonomous Multi-Cluster Remediation." More about his session: What if your on-call never had to touch a cluster again? This talk demonstrates an agentic AI system combining a Kubernetes MCP server layer across multiple clusters with an LLM orchestration loop to detect, diagnose and remediate issues autonomously. We cover the hub-and-spoke MCP architecture, agent guardrails, and real lessons from running this at scale. 🤖 See you soon!

Join us on Thursday, 5/28 for the Plural x Kubernetes May meetup 👋 ​Our guest speaker is @brunochauvet, Cloud Infrastructure Tech Lead at Rokt. His topic is "No More War Rooms: Kubernetes MCP Servers Agentic AI for Autonomous Multi-Cluster Remediation." 💡 🤖 What if your on-call never had to touch a cluster again? This talk demonstrates an agentic AI system combining a Kubernetes MCP server layer across multiple clusters with an LLM orchestration loop to detect, diagnose and remediate issues autonomously. We cover the hub-and-spoke MCP architecture, agent guardrails, and real lessons from running this at scale. 🗓️ ​Schedule: 6:00pm - door opens 6:30pm - intros (please arrive by this time!) 6:40pm - speaker programming 7:20pm - networking 8:00pm - event ends ✅ ​RSVP at: luma.com/g0ywmvfq See you there!

This week we hosted Plural x Kubernetes April meetup with our first returning guest speaker, @TheNJDevOpsGuy ! Key takeaways 👇 Michael's interactive session was on how to plug security holes by understanding the current standards, authentication at the system & user level, where AI gateway can help secure traffic, and how to specify tools exposed from MCP Servers with traffic policies. Quick hits: 1️⃣ OBO for all things agentic security 2️⃣ Agents need identities 3️⃣ Think about your line of communication (the gateway) 4️⃣ The majority of security is at the gateway level 5️⃣ Traces are let for end-to-end agentic health Our deep gratitude to Primary Venture Partners for generously hosting us in their event space 🙏 Mark your calendar for the next meetup on May 28 - invite to come :)

Join us on Tuesday, 4/28, for the Plural x Kubernetes April meetup 👋 We welcome back the first return speaker: ​Michael Levan (@TheNJDevOpsGuy), AI Architect at Solo and CNCF Ambassador. His topic is "Plugging Security Holes in LLMs and MCP Servers." 💡 🛠️ We will discuss how to plug security holes by understanding the current standards (stdio and streamable http), authentication at both the system and user level (jwt, oAuth, and OIDC), where an AI gateway can help secure traffic throughout the tunnel, and how to specify what tools should be exposed from MCP Servers with traffic policies. Bring your questions :) 🗓️ ​Schedule: 6:00pm - door opens 6:30pm - intros (please arrive by this time!) 6:40pm - speaker programming 7:20pm - networking 8:00pm - event ends ✅ RSVP at luma.com/sowljado See you there!

We're excited for #HumanX this week! Come by the Plural booth to get a demo and see Plural's always-on DevOps agents in action!

Another fantastic #Kubernetes #NYC meetup in the books! Key points from Marosha Afridi's session on the importance and know-how of identifying and fixing vulnerabilities: 🔍 The Reality Gap: Scanners report CVEs but stop there. The real challenge begins after the alert, where most practitioners have no structured path to actually remediate findings in a Kubernetes environment. 📝 Labels & Foundation Matter First: Before any remediation can be routed or tracked, workloads must be properly labeled (team, environment, criticality). Without this, you can't assign ownership or enforce SLAs. ❗ Package CVE ≠ Image Fix: Scanners flag a vulnerable package (e.g., openssl), but you can't just patch a package inside a container. You must identify the correct upstream image version, rebuild, retag, and redeploy , a process most orgs have no defined workflow for. 🛠️ Helm & GitOps Add Hidden Complexity: Most organizations deploy via Helm charts and GitOps pipelines. Updating a vulnerable image isn't a one-liner, it's a values change → PR → review → pipeline → verification cycle that can span days. 🌐 Multi-Cluster Attribution & Prioritization Beyond CVSS: The same CVE can appear across multiple clusters, namespaces, and teams. Raw CVSS scores are misleading. EPSS, reachability, and exploit availability must drive actual priority, not just the score. 🔄 Fix Verification Closes the Loop: Remediation isn't complete until you can prove the fix landed. Did the patched image actually deploy? Did the scanner re-run and confirm the CVE is gone? This final layer is where most programs fail silently. Anything else you took away from the session? Let us know in the replies👇 See you all next time~

📣 Last call for the March #Kubernetes NYC meetup happening tomorrow! Join us at 6pm to hear from Marosha Afridi, Senior Security Defensive Engineer at SAP. She will be giving a talk on "Stop Chasing Packages: Fixing Vulnerabilities the Container Way." ✅ RSVP at luma.com/9j2zs9sv 💬 About the Session: Today, container scanning tools are package centric, but organizations operate in an image centric world. Security tools tell us which package is vulnerable and what version to upgrade, but engineering teams don’t patch packages in running systems rather they rebuild and redeploy images. The missing capability is visibility into which image already includes the fix, reducing friction, lowering MTTR, and aligning security with how containers actually work. See you soon!

This is exactly what we're building @plural_sh. The core problem is that the on-the-ground reality is these tools are a zoo of duct tape, tribal knowledge, and elbow grease. We built a home-grown gitops engine to provide uniform execution which is now natively integrated with a full intelligence toolkit.

We're excited for #KubeCon next week! Swing by booth 684 to chat with our experts about AI-Native Kubernetes management. We'll see you there!

Join us Tuesday, 3/31 at 6pm for the March Kubernetes NYC meetup 👋 ✅ RSVP: luma.com/9j2zs9sv 🎙️Guest Speaker: Marosha Afridi, Senior Security Defensive Engineer at SAP 💡Topic: "Stop Chasing Packages: Fixing Vulnerabilities the Container Way" ​Session Abstract: Today, container scanning tools are package centric, but organizations operate in an image centric world. Security tools tell us which package is vulnerable and what version to upgrade, but engineering teams don’t patch packages in running systems rather they rebuild and redeploy images. The missing capability is visibility into which image already includes the fix, reducing friction, lowering MTTR, and aligning security with how containers actually work. ​We will have food and drinks during this event. Please arrive no later than 6:30pm so we can get started promptly. Hope to see you there!

For what it's worth, this is exactly what we're building against at plural.sh. The solve here is to remove terraform from a local dev loop and incorporate a full GitOps pr-based process around it, and Plural provides all the same AI sugar you love from Claude Code on top of that, ensuring you can move as fast as you want, but safely.

The obvious solution here is the ai should not be directly modifying a cloud account, but instead generating pull requests against a GitOps-style codebase to ultimately perform the change, a flow we're specializing in at Plural

February #Kubernetes #NYC meetup ✅ Key takeaways 👇 Avinash Sabat led an engaging presentation on Platform Engineering for Kubernetes, in front of a global audience (can we say that if there were international attendees??) We learned about: ⚙️ Programmable Control Plane: Kubernetes provides an extensible, API-driven core for orchestrating infrastructure and applications. 🛡️ Multi-Cluster Resilience: Distributing workloads across clusters enhances isolation, reduces blast radius, and enforces consistent governance. 🚦 Structured Abstractions: Platform Engineering builds high-level, opinionated layers on Kubernetes, simplifying developer workflows. 🧠 Reduced Cognitive Load: Platform abstractions empower developers by minimizing operational complexity while maintaining necessary flexibility. 🌐 GitOps at Scale: GitOps ensures declarative, automated, and auditable management of configurations and deployments across environments. Our deep gratitude to Resident Company Club for generously hosting us in your beautiful space 🙏 Know someone who'd be a good speaker or you've got a topic in mind? Send them our way~ The next NYC meetup will be at the end of March - let us know if you want an invite 👋

Join us Tuesday, 2/17 for the February #Kubernetes #NYC meetup 👋 ​ ✅ RSVP at: luma.com/r29heij9 Our guest speaker is Avinash Sabat, Cloud DevOps & Platform Engineer at Synechron and IEEE Senior Member! He will be doing a deep dive on platform engineering for Kubernetes at scale. 💡 ​Running Kubernetes at scale often exposes teams to more complexity than necessary, from manifests and namespaces to pipelines and policies. Avinash's talk will cover how platform engineering introduces higher-level abstractions on top of Kubernetes, enabling standardized, self-service workflows that balance governance, security, and team autonomy without slowing delivery. See you soon!

We are kicking off the year with the first #Kubernetes NYC meetup on Tuesday, 1/27! Michael Levan (@TheNJDevOpsGuy) is returning as a speaker to talk about #LLM and #MCP security with #agentgateway 💡 Space is limited, save your seat at luma.com/c2sv5uef 🗓️ Schedule: 6:00pm - door opens 6:30pm - intros (please arrive by this time) 6:40pm - speaker programming 7:20pm - networking 8:00pm - event ends See you there!