Introducing the next chapter of Redis Iris. Six new capabilities across retrieval, memory, routing, and messaging.
REDIS IRISAgents operate in an unreliable, messy world. Redis Iris is the unified, real-time engine that delivers relevant context so agents stay accurate, accountable, and efficient at scale.
Agents fail when they're missing context: the right information, at the right time, in the right form. But enterprise data is scattered across countless systems, making it stale, slow, and hard to navigate. The result is inaccurate answers, sluggish performance, and frustrated users.
The best outcomes happen when agents can follow connections between the available data and tools. Context turns both into a connected system agents can explore and action, not just query.
Agents keep preferences, decisions, and past outcomes that matter. They let go of what’s stale, superseded, or irrelevant. Remembering gives agents continuity, and forgetting keeps context relevant—and safe.
Every redundant LLM call, bloated prompt, and repeat lookup adds up cost and latency. Agents that retrieve only what’s relevant and reuse answers and procedures they’ve already generated cost less.
Context should get more relevant, personalized, and informed by past interactions. For agents to get smarter over time, they require built-in memory, learning, and statefulness.
One engine, six capabilities, endless agent possibilities.
Match the meaning behind the questions to the right resource, so you never recompute the same intent twice.
Increase accuracy and decrease inference cost with reusable answers and repeatable procedures.
Give agents intelligent short-term memory and persistent context across conversations in a single, fully customizable managed memory layer.
Every context source available via a single MCP endpoint means AI apps can retrieve the right facts fast.
Turn agent behavior into an explainable, queryable record of decisions and complete trails to diagnose failures, control costs, and prove what happened.
Conversations shouldn’t vanish when a network drops. Agent Mesh provides persistence and routing for agent streams, so agents pick up exactly where they left off.
Route every question by meaning, not wording. Intent Router sends each request to the right place to answer it, whether that's a remembered FAQ, an agent skill, a human, or an LLM, so agents skip the full recompute and respond faster for less.
Stop answering the same question twice. Agent Playbooks recognizes prompts that mean the same thing and reuses responses, cutting waste, while agents grow more accurate over time.
Deliver working memory and long-term recall in one place. Keep current agent conversations tight with active context, while persisting vital long-term pieces like user preferences, past decisions, and more across sessions and channels.
Give agents clean, schema-first paths through all business data, including customers, orders, tickets, and more. Agents spend less time guessing and more time reliably getting the right answers.
Turn agent behavior from a black box into a queryable system of record. Lens captures the full data trail behind every decision, so when an agent fails, gives a wrong answer, or burns through tokens, you can see exactly why.
Keep agents talking through dropped connections. Agent Mesh lets agents communicate over ephemeral streams, then buffers the stream and reconnects automatically when a connection drops, so agents pick up right where they left off.
We surveyed IT and AI infrastructure leaders on how they build, feed, and govern context for production agents. Nearly all of them say it's the deciding factor in whether AI works. Almost none of them have built the systems to prove it.
8 questions. 2 minutes. Recommendations for your team.
Memory is core to how agents improve over time, and teams we work with are realizing they need scalable infrastructure behind it. Together, Redis and LangSmith's Context Hub give customers a structured way to manage agent context across environments, connecting live operational data and retrieval into the same system that versions and evolves agent memory over time.
Everything you see in Paradigm is stored in Redis. It is the memory and orchestration layer that lets our agents work in real time, at a scale that wasn’t possible before.
Redis Agent Memory helps us store and reuse context across our coding agents in real time. We use it to capture critical engineering decisions, bug details, and development context as our agents work, so that information is available across the team instead of getting lost between sessions or tools. We also have a service reading from and writing to the same memory while monitoring our API, which lets us detect issues and report bugs into our ticketing system in real time. We’re excited to see what we can build next with Redis Iris.
See how Redis Iris keeps your agents accurate, accountable, and efficient. One runtime, six capabilities, endless agent possibilities.
