Do you test your AI agents before launch? 🤖 My colleague Jan-Felix Schmakeit shared 5 rules for evals: - Match graders to sandboxes - Write multi-step prompts - Avoid prompt-grader mismatch - Grade final outputs, not tool paths - Curate distinct user datasets Link👇 #AI #Evals #AgenticAI

Working on Google Cloud with AI agents? ☁️🤖 Install the latest google-cloud-developer plugin! 🚀 🧰 Core skills: gcloud, auth, onboarding 📚 Docs MCP server 🗂️ finding-google-skills Works with Antigravity, Claude Code, Codex, Gemini CLI. ⚡ Try and share feedback! 💬 Link below 👇 #GoogleCloud #AI

Level up your dev stack and build with Google Cloud AI! 🚀 Join me at #BuilderConnectPoland on Nov 3 in Warsaw. 🇵🇱 What to expect: • Engaging speakers 🎤 • Hands-on coding 💻 • Peer networking and 1on1 chats 🤝 • Swag and challenges 🎁 Link to register in the replies! 👇 #GoogleCloud #AI #Devs

Are you scared of quantum computers breaking your encryption and digital signatures? 🔐 Cloud KMS recently introduced a solution with quantum-safe signing: • ML-DSA (FIPS 204) • SLH-DSA (FIPS 205) • Fast external-µ signing Link in the replies! 👇 #CloudSecurity #Cryptography #DevOps

Fresh from the studio!

@kubernetespodcast.com · last mo.

Is cutting cloud spend enough to cut carbon? 🌿 In our latest episode, Niki Manoledaki unpacks Project Kepler, measuring real cloud energy consumption, architectural updates, and how AI is changing sustainability metrics. 📃 kubernetespodcast.com/episode/270-... 🔊 kubernetespodcast.com/subscribe

👀 Sneak peek behind the scenes of Google Agent Skills! In my new article, I share how we build, test, and scale AI skills at Google: • Standardized layout • Remote MCP tools for IAM • Automated CI/CD checks • 2x2 evaluation matrix Full article link in replies 👇 #GoogleCloud #AIAgents #AgentSkills

Nowy polski model embeddingowy PolDense-1B wysoko w rankingu PIRB! 🚀 Linki wrzucam w odpowiedzi. Czy ktoś z Was testował go już w swoich RAGach? Jakie macie wrażenia? Podzielcie się opiniami! 👇

My crazy Italian friend is running 20 subagents in parallel on the same codebase. What can go wrong?

Riccardo C@palladius.bsky.social · 2mo ago

Oulala! 🦖 I Tokenmaxxed 20 subagents to fix all my GitHub issues at once on Rails! Read the article here: medium.com/@palladiusbo... Or steal my prompt & agent skill directly: github.com/palladius/ge... #Rails #AI #GoogleCloud #Ruby

The (no longer) missing multi-agent pattern: triggering dynamic workflows from an agent. 🧩🤖 Google ADK 2.4.0 makes dynamic workflows first-class tools🛠️ Full article with Python code examples 👉 dev.to/googleai/the... #AI #GoogleADK #SoftwareEngineering #ADK #AgenticAI #Agents

The (no longer) missing multi-agent pattern: triggering dynamic workflows from an agent

When building multi-agent systems, rigid state graphs quickly fall apart in the face of dynamic user...

dev.to

Casey analyzes and deconstructs the viral case of a coffee shop that was run by AI agents as an experiment. How can we avoid similar mistakes?

Casey West@caseywest.bsky.social · 2mo ago

They swapped the model mid-experiment. The café failed the same way. So the model wasn't the problem. It shipped with no spend cap and no approval gate—the seatbelts you install on day one. The proof and the five fixes: www.linkedin.com/pulse/ai-did...

Today Gemini CLI transitioned to Antigravity CLI 🚀 Read my guide: "Building an agentic PR reviewer with Antigravity SDK". Automate your PR reviews and avoid the AI orchestration tax. Let an agent find bugs before you review the code! 🐞 Link below 👇 #Antigravity #CodeReview #AI #DevOps #GeminiCLI

I will be openning the Developer track at Google Cloud Summit in Prague on 9 June 🇨🇿 Join me for technical content and an opportunity to discuss your ideas. Agenda and registration link below 👇 #GoogleCloud #CloudSummit #GoogleCloudSummit

Yesterday's Google I/O was filled with announcements 🚀 I am particularly excited about the Antigravity 2.0 release 🤖 , which transforms it into a standalone desktop app alongside new IDE, CLI and SDK components. Read more at: blog.google/innovation-a... #agy #googlecloud #googleio #antigravity2

Building the agentic future: Developer highlights from I/O 2026

We're introducing new tools to take your ideas from a prompt to a production-ready application, including updates to Google Antigravity, an enhanced Gemini API and nativ…

blog.google

How do you connect AI to physical hardware? 💻🔒 Sweets Vault uses the ADK and Gemini API to verify homework and unlock real reward drawers. Read the guide to learn about multimodal verification, state management, and HW integration. Link in comments! 👇 #Gemini #AgenticAI

Kubernetes Podcast is back with a bang!

@kubernetespodcast.com · 4mo ago

This episode we speak to Lucy Sweet. Staff Engineer at Uber Engineering and lead for the Kubernetes (Official) Node Lifecycle WG. Lucy shares Uber's 5 Millions cores on #k8s story at scale and why it's always DNS 😃 📃 kubernetespodcast.com/episode/266-... 🔊 kubernetespodcast.com/subscribe

🚀 Spoke at AI Tinkerers Wrocław on May 6. I demoed Sweets Vault. It is a gamified agentic homework validation tool for kids. It uses a RPi and a LED display. Great to meet other presenters and watch their demos. Thanks to the organizers for a stellar event. #AITinkerers #AI

See my latest blog for a behind the curtain look at how I used Gemini CLI and Conductor to manage a full RAG migration. - Spec-driven development 📝 - AI-driven TDD 🧪 - Infrastructure as Code ☁️ Scalable engineering is here. Link in the replies! ⬇️ #AI #Coding #GoogleCloud

Stop choosing between upgrading AI models and search availability. Use shadow deployments to keep your RAG system live. 🚀 Link to blog in comments 👇 #AI #VectorSearch #RAG #CloudRun #GoogleCloud

Building scalable RAG backend is easier than you think. 🛠️ New guide on using: 🔹 Cloud Run Jobs for parallel embeddings 🔹 AlloyDB + pgvector for storage 🔹 Vertex AI & Terraform Part 1 of my series is live! Link in the replies. 👇 #RAG #GoogleCloud #GenAI #PostgreSQL

Building production agents is hard. Join our meetup at Next 26 on April 22 at 4:00 PM 🗓️ No slides. Peer talks on reliability and orchestration. 🤖 See comments for session link 👇 You can also find me at the Data Cloud booth to say hi 👋 #GoogleCloudNext #AIAgents #GenAI

Big day for open-weight models. Gemma 4 is out 🚀 I am eager to hear about all those local use cases you're working on. Share in the comments 👇

Google Open Source@opensource.google · 6mo ago

With today's Gemma4 announcement, you might want to try the model yourself and not know where to start. Well how about an AIventure where you learn about vibe-coding, agents and gen-ai with Gemma4 specifically! developers.google.com/solutions/le...

Cloud Run Jobs or Cloud Batch? If you're running offline processing on Google Cloud, the choice isn't always obvious. While both services are built for run-to-completion tasks, they serve very different needs. Let me help you choose with my guide: medium.com/p/8590a8e3a3b1

Cloud Run Jobs vs. Cloud Batch: Choosing Your Engine for Run-to-Completion Workloads

Google Cloud offers plenty of different products and services, some of which seem to be covering overlapping needs. There are multiple…

medium.com

0% High-level talk. 100% Engineering Velocity. 🚀More than 300 devs brought the energy to Google Cloud Builders Day Berlin on March 17! Next up: Google Cloud Summit DACH | Frankfurt | June 9–10 Register here → cloud.google.com/events/summit-dach-2026-developer #GoogleCloud #BuildersDay

Stop hand-holding your AI agents! 🤖 Run Gemini CLI in "yolo mode" safely with a custom Docker sandbox and limited permissions. 🛡️ Link in the first comment! 👇 #GeminiCLI #GoogleCloud #AI #SoftwareDevelopment #AgenticAI

GKE Private Cluster is a great way to keep your workloads more secure.. However, no Internet means no Docker Hub and no Hugging Face! So how do you deploy an LLM inference service on a Private Cluster? Have a read and find out: medium.com/p/70a23cc9c315

Inference on GKE Private Clusters

Deploying an inference service on your GKE private cluster using Artifact Registry, Cloud Storage and Persistent Disks.

medium.com

Coding with a Google Cloud customer in 🇸🇪Stockholm this week. We simplified the architecture and used Gemini CLI with Antigravity to iterate fast🛠️ #GoogleCloud #Stockholm #Cloud #AI

Join my workshop at Builders Day Berlin on March 17. Build MCP servers, connect Gemini CLI, and deploy AI Agents to Cloud Run. Get hands-on experience with ADK. 🚀 Registration link in the comments 👇 #GoogleCloud #CloudRun #BerlinTech #AI #Developers

Run EmbeddingGemma on Cloud Run with L4 GPUs! 🚀 New lab shows you how to: - Package everything in Docker 🐳 - Deploy to serverless GPUs ⚡ - Build a semantic search app 🔍 Scale your RAG systems without managing infra. Link in comments👇 #CloudRun #Gemma #AI #Serverless

Check out this infographic! 📊 No UX experience? No problem. I dropped a new guide on grabbing a #GoogleCloud Free Trial to enable Nano Banana Pro and start playing with high-fidelity visuals. 🍌 Link in the comments👇 #Gemini3 #NanoBananaPro #WinterOlimpics

Stop babysitting your AI agents. 🍼 The "Ralph Wiggum" loop is all about autonomous, self-correcting Agent that doesn't stop until the job is done. I’m using Antigravity to handle grind while I focus on other tasks. Full guide below! 🚀 #AI #Antigravity #Coding #RalphWiggum