Redis

Backed by standard Redis data structures (HASH and SET keys). No custom module required — it works with any Redis instance. Point your agent here: github.com/rowantrollope/… and tell it to migrate its memory directory to redis-fs.

Today, we launched the Redis Partner Network. Companies aren’t just experimenting with AI anymore. They’re rebuilding apps around fast data and performance. Shifts like that don’t happen alone. They take partners who know how to deliver, help customers move from idea to production, and stay focused on outcomes. We built this program to make it easier to grow together, work closer, and drive real results for customers. Excited to build what’s next with an incredible group of partners. Take a look and get started with us today. redis.io/partners/

What if you could manage your 𝗥𝗲𝗱𝗶𝘀 𝗖𝗹𝗼𝘂𝗱 resources using plain English? Or any other language? I'm excited to share that @Redisinc has an MCP Server for Redis Cloud now, and it is super easy to get started with. You can get it started with 3️⃣ simple steps. 🧵

Paradigm can take a list of target companies, search their public web presence and GitHub footprint, and return evidence of Redis adoption: a careers page hiring for engineers with Redis experience, a public repo using a Redis client, or docs referencing caching/session infrastructure built on Redis. In <1 minute. And set it up to monitor into the future or sync with our CRM. Below you can see how @paradigmai identified Redis usage at top coding agent companies.

Agents love the filesystem with MD files. But when you want to share your filesystem across multiple agents its tricky Check out redis-fs a POSIX compliant filesystem backed by Redis. Mounts as FUSE on Linux and NFS on Mac. github.com/rowantrollope/…

Yesterday, @bsbodden and I delivered our Designing Multi Agent Systems with Spring AI hands-on lab at JavaOne in San Francisco. More than 35 people showed up and it was the lab with most registrations of the day. 🙏🙌☕️

@Bikash__Shaw We think Redis is awesome too. If you're using Valkey, though, you're not using Redis. Here's a quick comparison with a link to sign up for Redis Open Source here: redis.io/compare/valkey/

Just added Redis caching (Valkey in a Docker container) to my Urbanease backend. Response time went from ~4000ms -> 2ms🤯 Still figuring out more handler to use this with proper invalidation time to avoid stale issues. @Redisinc is awesome!!

A lot of developers still think of Redis as "just a cache". That undersells it, especially for AI integrations. What makes Redis useful in AI is the same thing that made it useful in distributed systems for years: It protects your hot path. Once you add AI features to an application, new problems show up very quickly: - Every request wants more context - Some requests are semantically similar - Repeated model calls get expensive - Latency becomes much harder to control - Your primary database starts handling work it was never meant to optimize for This is where Redis gets interesting. Redis can already sit on the hot path of a system with very low-latency reads and writes. On top of that, Redis now supports vector search, metadata filtering, semantic caching, and AI memory patterns. That makes it useful for much more than caching: - Retrieval for RAG - Semantic search - Conversation/session memory - Semantically similar response reuse - Low-latency context access before a model call If you’re building in .NET, this matters. That’s also why RedisVL caught my attention. RedisVL is Redis’s Python library for AI workloads. It gives you higher-level building blocks for things like vector search, semantic caching, and memory, instead of forcing you to stitch everything together from scratch. Yes, RedisVL itself is Python-first. But the bigger idea applies well beyond Python: AI systems need more than model access. They need low-latency infrastructure around the model. I think more .NET teams are going to run into this over the next year. If you want to explore how Redis fits into AI systems, start here: fandf.co/4l3ZXBe A huge thank you to @Redisinc for collaborating with me on this post.

At @NVIDIAGTC, the conversation isn’t just about agents and LLMs, it’s also about context. Getting the right context to the model in real time is what separates production-grade agents from the ones that fall apart in practice. We broke it all down in a guide about context engineering and agent memory. Come see how devs are using Redis as a real-time context engine: redis.io/gtc

You may have heard the rumors at #GTC. Our boss was upset we didn’t get a booth, but at least we let you know we are the real-time context engine for AI. See how devs are building with Redis: redis.io/gtc

More in this guide: redis.io/blog/llm-conte…

“Context window exceeded.” If you build LLM apps, you’ve seen this before. Context windows control how much text a model can process at once. They impact prompt design, RAG architecture, model selection, cost, and performance. If you’re building production AI systems, you need to understand them. Start with the fundamentals: • What context windows are • Why they exist • How they work • How to choose models strategically

🤖 Better AI is not just about better models. 🎙️ @mvizard speaks with Redis Context Engine Lead Simba Khadder about why context engines matter and why giving AI agents direct text-to-SQL access is a serious risk. Watch full interview here: techstrong.tv/videos/ai-lead…

Here’s what support actually means in production: redis.io/blog/redis-vs-…

Reachability and accountability aren’t the same thing. The different matters in production. With ElastiCache, support and SLAs are tied to managed service availability. If the infrastructure is reachable and meets uptime targets, the commitment is met. The focus is on the platform layer. With Redis, support is defined around Redis itself. We provide Redis-owned SLAs, 24x7 support, defined response times, formal escalation, and lifecycle management—across clouds and self-managed environments. The difference becomes clear when something breaks and the question shifts from “Is it up?” to “Is it behaving correctly?” Make Redis foundational to your system for support that is part of your reliability strategy, not just a line item.

@debasishg Indeed, I wrote this blog post: antirez.com/news/156 It was hardly considered by anybody ;D so very happy you are interested. Thanks.

Agents are shipping everywhere with @LangChain and Redis. Those that survive production aren't just prompting better—they’re engineering better context. The problem? “Context engineering” gets tossed around like it’s obvious. It’s not. It’s a skill you build. That’s why we launched our free “Context Engineering with Redis & LangChain” lab on Redis University. You’ll build a course advisor agent step-by-step, layering system context, RAG that actually improves answers, persistent memory, and user intent—using Redis for vector search and agent memory, all orchestrated with LangChain. If context engineering has felt like a buzzword, this makes it real. Get started on Redis University: university.redis.io/course/vsgabnb…

Register: redis.io/resources/vide…

We’re live today with @Redisinc! Join Design AI Agent Memory for Scale to hear how we think about building durable AI agent memory for real-world systems. We’ll share what actually matters when building for scale. Register from the link below!

When network latency, slow queries, and server saturation stack up, response times climb fast. Much of that overhead comes from repeated disk I/O and database round trips. The right caching layer removes it. Redis serves frequently accessed data from memory, cutting round trips and reducing tail latency at scale. Measure your baseline. Focus on p95 and p99. Identify the bottleneck. Cache what’s hot. Here’s how to take control of your apps response time: redis.io/blog/app-respo…