Managed Redis cache is a turnkey "Shared State" add-on for Kong Dedicated Cloud Gateways. It is designed to combine the performance of an in-memory data store with the simplicity of a SaaS product.

When you spin up a Dedicated Cloud Gateway in Kong Konnect, you can now toggle on Managed Redis with a single API call or a simple line in your Terraform provider. Kong handles the rest—provisioning, patching, scaling, and ensuring the Redis instance is co-located with your dataplane for sub-millisecond latency.

Key Capabilities

The managed Redis cache makes it easy for organizations to transform the DCGW from a stateless proxy into an intelligent, state-aware platform designed for the AI era:

• Accelerating Agentic AI: Seamlessly enable AI Rate Limiting and AI Quota plugins to govern AI agent consumption across models (e.g., OpenAI, Anthropic) without managing a separate database.
• Seamless Multicloud State: Delivers a dedicated, regionally scoped cache layer in each cloud environment, optimizing performance and maintaining consistent rate limits and caching behavior wherever your gateway runs.
• Native Performance for Global Users: Redis instances are co-located with the dataplane in the customer’s cloud region, ensuring optimal performance and minimal round-trip latency.
• Zero-Ops Management: Removes the need for customers to provision, scale, or secure Redis clusters. Configuration is handled entirely through Konnect APIs, Terraform and the UI (coming soon!).

As enterprises integrate LLMs into their products, AI and API gateways have become the "AI Firewall." Agentic AI models can consume tokens at an unpredictable rate, and without strict governance and semantic caching, costs can spiral out of control.

With a managed state layer, you can achieve the following:

• Enforce AI Quotas Globally: Set a hard cap on the number of tokens an "Agent" can consume across OpenAI, Anthropic, and Gemini, even if those requests are distributed across multiple gateway instances within a region.
• Semantic Caching: Cache expensive LLM responses to serve subsequent identical prompts instantly, slashing both latency and API costs.
• Context Management: Maintain the state of long-running AI sessions.

Without the capabilities above, governing AI can be a manual, reactive process. With it, governance becomes an automated, proactive shield.

The "Multi" in multicloud often leads to "Multi-headaches." Organizations can find themselves managing different sets of tools for AWS and Azure. Kong Dedicated Cloud Gateway provides a consistent operational experience.

Whether your Dedicated Cloud Gateway is running on AWS or Azure, the way you interact with it remains the same. This abstraction allows platform teams to build "Global Policies." You can write a single Terraform module that deploys a gateway with Managed Redis to five different regions across two cloud providers, ensuring that your caching and rate-limiting logic is identical everywhere.

This eliminates the "cloud-specific" configuration drift that can lead to security vulnerabilities or performance inconsistencies.

Managed Redis isn't just about AI and multicloud; it’s about making every API interaction faster and more reliable.

1. Advanced Traffic Control

Standard rate limiting is often "node-local," meaning each gateway instance counts requests independently. In a distributed environment, this can lead to inaccuracies and unintended consequences. With a synchronized rate limit policy within a region, all gateway instances are subject to precise and consistent rate limiting.

2. Shield Fragile Legacy Backends

Many legacy systems weren't built for the scale of modern mobile apps. By enabling Proxy Caching, you can shield your fragile backends from traffic spikes. The gateway serves the cached response directly from Redis, reducing the load on your core systems and improving the end-user experience.

3. Improving Reliability (Stale-While-Revalidate)

The gateway cache can provide a "safety net" if your backend goes down. E.g., if the backend returns a 5xx error, the gateway can be configured to serve the last known good version of the data from the cache. The user sees a slightly "stale" page rather than a "Service Unavailable" error.

In an era defined by architectural modernization and the surge of AI, the need for secure, resilient, and high-velocity API infrastructure is at an all-time high. Kong Cloud Gateways has met this challenge head-on over the last year, broadening its multi-cloud footprint and delivering critical enhancements in networking, uptime, and enterprise-grade security. Highlights have included expansion into Google Cloud, SOC 2 and PCI DSS compliance, and achievement of 99.99% SLA.

We are committed to making Kong Konnect the most powerful multicloud API platform on the planet. Managed Redis for DCGW is currently available across AWS and Azure.

In an era where "State is Great," don't let your infrastructure hold your applications back. Whether you are governing autonomous AI agents or simply trying to provide a consistent rate-limiting experience across multiple clouds, Managed Redis is the foundation you’ve been waiting for.