Redis Iris Empowers AI Agents with Memory and Context for Efficient Production Deployment
On November 6, 2026, Redis announced the official launch of Redis Iris, its real-time context engine designed specifically for AI agents. This latest version offers significant advancements, providing a suite of six customizable features that promote reliability and cost efficiency for AI agents in production environments. These features include Intent Router, Agent Playbooks, Agent Memory, Context Gateway, Lens, and Agent Mesh.
The emergence of AI models for enterprise use has accelerated, yet many teams struggle to deploy agents that function reliably in operational settings. A common challenge lies in ensuring agents possess appropriate context, which must be readily available and accurately retained in memory. Inefficiencies often surface from overloaded prompts, redundant calls to large language models (LLMs), and fragile custom integrations that can lead to substantial token costs outweighing generated value.
"Developers shouldn't require multiple infrastructure products for agents to function properly. Iris consolidates these essential capabilities into a single engine, allowing developers to create efficient production-ready agents," stated Rowan Trollope, CEO of Redis. He emphasized that agents need foundational elements similar to applications and microservices, such as real-time application data, runtime state, and service communication, alongside new functionalities for memory management, context handling, and routing.
Redis Iris delivers production-ready capabilities for context, memory, and messaging right from the start, allowing teams to tailor each feature to their agents. The functionalities include:
Together, these functionalities enhance agents' capacity to remember pertinent information, respond accurately, and operate transparently while maximizing results per token. Built on the Redis technology already familiar to many teams, Iris allows organizations to streamline integrations typically required for efficient information retrieval, memory management, routing, and messaging.
In production, agents face unpredictable user inputs, multi-channel communication, and task transitions, which can all become potential failure points. Redis Iris operates on four core principles: contextual data must be navigable rather than siloed; memory must be adaptive and selective; context use should optimize cost efficiency; and it should grow as agents learn from past interactions.
Since its initial release in May 2026, Redis Iris has already been harnessed in operational settings by forward-thinking companies. For instance, San Francisco-based Paradigm, recognized by Fortune as a game-changer in replacing traditional spreadsheets with swarms of AI agents, has built its entire platform on Redis. June Lee, CTO of Paradigm, noted that "Everything we see at Paradigm is stored in Redis. It serves as the memory and orchestration layer that enables our agents to operate in real time at an unprecedented scale."
As of today, Redis Iris is available with Intent Router, Agent Playbooks, Agent Memory, and Context Gateway in general release, while Lens and Agent Mesh are still in beta testing. For more details, visit redis.io/iris.
To explore Redis Iris further, Redis is hosting a webinar on November 12 for the Americas and November 17 for the EMEA region, with registration open at redis.io/resources/videos/redis-iris-full-spectrum.
Redis is recognized as the fastest data platform globally. Originating from open-source roots in 2011, Redis has evolved into a leader in caching solutions, aiding over 10,000 clients in building, scaling, and deploying the applications that power our world. With databases available both in the cloud and on-premises for caching and generative AI use, Redis sets a new standard for application speed. With a presence across San Francisco, Austin, London, Tel Aviv, and Singapore, Redis is internationally acknowledged for leading the development of high-performance applications. For more information, visit redis.io.
The emergence of AI models for enterprise use has accelerated, yet many teams struggle to deploy agents that function reliably in operational settings. A common challenge lies in ensuring agents possess appropriate context, which must be readily available and accurately retained in memory. Inefficiencies often surface from overloaded prompts, redundant calls to large language models (LLMs), and fragile custom integrations that can lead to substantial token costs outweighing generated value.
"Developers shouldn't require multiple infrastructure products for agents to function properly. Iris consolidates these essential capabilities into a single engine, allowing developers to create efficient production-ready agents," stated Rowan Trollope, CEO of Redis. He emphasized that agents need foundational elements similar to applications and microservices, such as real-time application data, runtime state, and service communication, alongside new functionalities for memory management, context handling, and routing.
Six Features, One Engine
Redis Iris delivers production-ready capabilities for context, memory, and messaging right from the start, allowing teams to tailor each feature to their agents. The functionalities include:
- - Intent Router (GA): This feature interprets the intention behind a request rather than its wording, directing it to the appropriate response, skill, person, or model. It also enables routing to agents and models, applies safeguards, and uses fuzzy conditional statements to inform decision-making beyond binary true/false.
- - Agent Playbooks (GA): This allows teams to identify stable responses that should be reused versus those that require updated data, enhancing efficiency with each execution by reducing resolution times and token expenditures.
- - Agent Memory (GA): Through MemoryETL, this feature generates memory elements from user actions and semi-structured data, organizing them hierarchically for better control. It also provides inspection and auditing capabilities while ensuring privacy by allowing the forgetfulness of unnecessary information, such as personally identifiable details.
- - Context Gateway (GA): Acts as a single access point for agents to connected context sources, adhering to governance protocols. Teams can utilize their own MCP servers, cache tools that resolve specific issues to minimize token usage, and generate reusable outputs to maintain session consistency.
- - Lens (beta): This tool tracks agent behaviors and offers traceability by linking each outcome to associated skills or FAQs, accompanying memory elements, and originating sessions, providing auditable evidence and insight into performance metrics like task completion rates and latency.
- - Agent Mesh (beta): Facilitates ephemeral communication between agents, managing buffering and reconnection automatically during connectivity disruptions. Persistent SSE streams maintain response flows, even amidst interruptions, while the A2A Comms protocol creates reliable channels for task delegation and inter-agent transfer.
Together, these functionalities enhance agents' capacity to remember pertinent information, respond accurately, and operate transparently while maximizing results per token. Built on the Redis technology already familiar to many teams, Iris allows organizations to streamline integrations typically required for efficient information retrieval, memory management, routing, and messaging.
Designed for Real-World Complexity
In production, agents face unpredictable user inputs, multi-channel communication, and task transitions, which can all become potential failure points. Redis Iris operates on four core principles: contextual data must be navigable rather than siloed; memory must be adaptive and selective; context use should optimize cost efficiency; and it should grow as agents learn from past interactions.
Since its initial release in May 2026, Redis Iris has already been harnessed in operational settings by forward-thinking companies. For instance, San Francisco-based Paradigm, recognized by Fortune as a game-changer in replacing traditional spreadsheets with swarms of AI agents, has built its entire platform on Redis. June Lee, CTO of Paradigm, noted that "Everything we see at Paradigm is stored in Redis. It serves as the memory and orchestration layer that enables our agents to operate in real time at an unprecedented scale."
Availability and Resources
As of today, Redis Iris is available with Intent Router, Agent Playbooks, Agent Memory, and Context Gateway in general release, while Lens and Agent Mesh are still in beta testing. For more details, visit redis.io/iris.
To explore Redis Iris further, Redis is hosting a webinar on November 12 for the Americas and November 17 for the EMEA region, with registration open at redis.io/resources/videos/redis-iris-full-spectrum.
About Redis
Redis is recognized as the fastest data platform globally. Originating from open-source roots in 2011, Redis has evolved into a leader in caching solutions, aiding over 10,000 clients in building, scaling, and deploying the applications that power our world. With databases available both in the cloud and on-premises for caching and generative AI use, Redis sets a new standard for application speed. With a presence across San Francisco, Austin, London, Tel Aviv, and Singapore, Redis is internationally acknowledged for leading the development of high-performance applications. For more information, visit redis.io.
Topics Consumer Technology）
【About Using Articles】
You can freely use the title and article content by linking to the page where the article is posted.
※ Images cannot be used.
【About Links】
Links are free to use.