Operations | Monitoring | ITSM | DevOps | Cloud

TL;DR OpenSearch 3.6 makes agentic search production-ready, with the AI-powered Launchpad provisioning full search apps in minutes and faster default vector search, and it's the first LTS release, bringing 18+ months of guaranteed support, SBOMs, and an upstream-first commitment (every fix goes back to the main project) so teams get fast-moving open source and a stable, supported platform at once.

Kafka is brilliant at moving events around, but sooner or later someone wants to actually query those events, perhaps aggregations, dashboards, or ad-hoc analytics over billions of rows. That is where ClickHouse comes in. It's the option for when stream processing is more than you need, but warehouse query latency is more than you'll tolerate.

I spend a good amount of my time deploying Aiven services for demos and examples. Traditionally the tools I reach for are: If I’m writing a program, I may also look to the Aiven API, perhaps using curl at the command line or in a shell script, or perhaps with direct HTTP requests in a Python program. The API is how the console and the CLI tool talk to Aiven, but I generally find that too low level to be comfortable, and I always have to look up how to pass in the Aiven user token.

You're building with Claude or Cursor, and you need to know what's actually happening on your Kafka cluster. Your AI assistant knows Apache Kafka in the abstract, but not your topics, your retention, or that a consumer group has been slipping since this morning. So you leave the editor and go digging through logs, a CLI, and a few dashboards, correlating by hand to answer questions like: The Aiven MCP (EA) turns each of those into a sentence you type where you already work.

This Spring, the Apache Kafka community released version 4.2 with a “production-ready” Share Group feature. Also known as a “Kafka queue” people were eager to see this feature because it introduced elastic consumer scaling, individual message acknowledgments, and built-in "poison pill" handling; similar to what you'd find in traditional message brokers like RabbitMQ and ActiveMQ.

Benchmarks keep showing that picking the right LLM is hard. The easy answer is "just use the most powerful one." That works, but it is pricey. A small, cheap, or local model can handle many simple requests just as well as a frontier model, for a fraction of the cost. That is what semantic routing is for. Use middleware that looks at an incoming request and decides which model should answer it.

I’ve been active in technical communities like PyData and codebar across the UK for nearly a decade now and owe much of my career to meeting and learning from cool people at events. Now, as someone with a professional interest in community organising, I spend a lot of my time looking for events to sponsor or speak at. But whilst finding technical events and communities can be tricky, I think I’ve found a solution using Kafka that might be useful for you too.

You've felt it. You're deep in a flow state with Claude or Cursor, building the next great thing, and then you hit the wall. Time to leave your editor, open a browser, click through a console, copy a connection string, paste it back, and pray you didn't fumble a character. The vibe is gone. What if your AI agent could just... do it? Deploy the database. Create the Kafka topic. Ship the app. All without you ever leaving the conversation. Today, that's real.

Aiven for Valkey includes the Valkey Search module setup and ready to go. Here's what that looks like in practice: a small online shop adding real search on top of the cache it's already running. Needle & Yarn sells the yarn you crochet with (skeins) and the design patterns you crochet from. Like a lot of e-commerce backends, it already runs Valkey as a product cache, with each product stored as a Hash for hot-path performance.

In this article we’ll show how we built an image search demo in Aiven Apps. The demo uses the CLIP Large Multimodal Model (LMM) to turn a user’s text prompts into a vector that can be compared with the precomputed vectors for a corpus of images, allowing the user to find images based on their text. While in this example the LMM input (the text prompt) is coming from the user, the principle is the same as for an internally generated query.