Giorgi

🔥 Say hello to Pyroscope 2.0! This is a ground-up rearchitecture of our open source continuous profiling database — designed to be more cost-effective at scale, simpler to operate, and with native #OTLP profiling support built in. grafana.com/blog/pyroscope…

🛡️ OpenAI Releases 5-Point Action Plan to Strengthen AI-Powered Cyber Defense Source: cybersecuritynews.com/openai-5-point… OpenAI has published a comprehensive cybersecurity action plan outlining a five-pillar strategy to equip trusted defenders with advanced AI capabilities while preventing adversarial misuse. Artificial intelligence is fundamentally reshaping the cybersecurity landscape and not just for defenders. Malicious actors are already leveraging AI to improve phishing campaigns, automate reconnaissance, accelerate malware development, evade detection, and scale cyber operations at unprecedented speed. #cybersecuritynews #OpenAI

📣 We're open sourcing o11y-bench, a benchmark for evaluating AI agents on observability workflows. Built on Harbor, it runs agents against a Grafana stack w/ access to our MCP server & grades them on a set of observability tasks within that environment. grafana.com/blog/o11y-benc…

🚨 BREAKING: Wiz Research discovered Remote Code Execution on GitHub.com with a single git push The flaw in @github allowed unauthorized access to millions of repositories belonging to other users and organizations 🤯

What if your deployments could fix themselves before users even notice? 🎥 Panel: From Signals To Safety Nets: Self-Healing Progressive Delivery With Argo Rollouts with Anastasiia Gubska (Chase UK), Julia Furst Morgado (@dash0hq) & Joe Fuller (Buoyant) Straight from ArgoCon Amsterdam 🇳🇱, this one dives into signal-driven, self-healing rollouts. You’ll see how teams: - Detect issues early and auto-rollback - Pause + scale based on real-time signals - Build closed-loop, resilient delivery systems Worth a watch (or rewatch) if you’re thinking about smarter releases. ▶️ Watch here:  youtube.com/watch?v=VXYU0o…

Keep your systems resilient. 🛠️ Learn the latest in logging, tracing, and monitoring at Observability Summit North America. Register by May 4 to save $200 on your Minneapolis pass! 🎟️ Get your ticket: bit.ly/3PmWO41 #O11ySummit

5 years of KCD Czech & Slovak! The community is heading back to Prague on May 21-22 to talk shop on all things #CloudNative. No fluff—just real-world usage from the people building and maintaining these stacks every day. Whether you're into Platform Engineering, Wasm, or scaling K8s, see you there. 🇨🇿🇸🇰 #KCDCZSK #Kubernetes #CNCF

MICROSOFT JUST DROPPED A 4B MODEL THAT TURNS ONE IMAGE INTO A FULLY TEXTURED 3D ASSET IN SECONDS

@morganlinton Eez noeone listens to :/

I’ve said it before and I’ll say it again, MCP is not secure, by design.

Tim Cook steps down as Apple’s CEO

tasted it, have some issues to cut video, but prolly worth to test build refresh.usectl.com to throw your anger away

This is how grok see us

Claude just dropped the new terminal + UI update. had to test it. gave it one prompt. it built a full Lovable clone. deployed it on usectl. the whole thing took 12 minutes. then i gave the screen recording to Claude Cowork and it cut a 2 min video from it. prerequisites: GitHub CLI and usectl CLI. thats literally it. now heres the thing. Lovable just shipped their payment integration as a feature. cool. but with Claude i can just say "add Stripe checkout to this project" and its done. easy onboarding, real Stripe SDK, full control. no platform middleman taking a cut of your payments. we're at the point where an IDE and usectl is all you need to build, deploy and monetize a real product. from idea to live app with payments in one afternoon. this is not a demo. this is Tuesday. @karpathy @Lovable @claude_code @claudeai

@kritikakodes Photoshop

Guess the stack?🤔

Kubernetes question that separates the YAML copypasters from the ones who’ve actually been burned: etcd is the ‘single source of truth’ for your whole cluster… yet 90% of teams only remember it exists when the cluster is already on fire. Kubelet? The quiet agent on every node that’s actually keeping your pods alive (or secretly killing them at 3am). Which one has caused you more production pain in 2026? Drop your war stories below — no theory, only scars 👇 #Kubernetes #K8s #etcd #kubelet"

They're just a spot where you can slap on random extra info about your pod, deployment, or whatever object you're creating. Think of them as sticky notes you attach to the thing — stuff that doesn't affect how Kubernetes itself runs or selects the object, but other tools and systems can read it and do their thing.

Kubernetes question In a Kubernetes manifest file, what is the role of annotations?

Just wrapped up an amazing few days at KubeCon + CloudNativeCon Europe in Amsterdam! 🌟 Incredible talks, great people, and so much energy in the cloud native community. What a fantastic experience! #KubeCon #CloudNative #CNCF #Amsterdam

What if your repo was the AI agent? GitAgent makes that real. • Git-native AI agent standard • Your repo = memory, skills, identity • Works across OpenAI, Claude, CrewAI, more • Versioned, reviewable, reproducible agents • Run agents straight from GitHub

🚨: For the first time, we are observing plants breathe in real time.

RAG vs. Graph RAG, explained visually! RAG has many issues. For instance, imagine you want to summarize a biography, and each chapter of the document covers a specific accomplishment of a person (P). This is difficult with naive RAG since it only retrieves the top-k relevant chunks, but this task needs the full context. Graph RAG solves this. The following visual depicts how it differs from naive RAG. The core idea is to: - Create a graph (entities & relationships) from documents. - Traverse the graph during retrieval to fetch context. - Pass the context to the LLM to get a response. Let's see how Graph RAG solves the above problem. First, a system (typically an LLM) will create a graph from documents. This graph will have a subgraph for the person (P) where each accomplishment is one hop away from the entity node of P. During summarization, the system can do a graph traversal to fetch all the relevant context related to P's accomplishments. The entire context will help the LLM produce a complete answer, while naive RAG won't. Graph RAG systems are also better than naive RAG systems because LLMs are inherently adept at reasoning with structured data. 👉 Over to you: Have you used Graph RAG in production? ____ Find me → @_avichawla Every day, I share tutorials and insights on DS, ML, LLMs, and RAGs.