Real-time context: keeping agent inputs fresh on every step
Stale context makes AI agents fail even when retrieval works. Learn how to keep agent inputs fresh at every step with a real-time data layer.
Redis Blog
@redis-blog
Publisher
61
Posts
Posts by Redis Blog
CEO Rowan Trollope’s organizational announcement to Redis employees
Rowan Trollope shares an organizational change at Redis and explains why the company is realigning to focus on the AI and agent era ahead.
Agent interoperability: a complete explainer
Learn how MCP, A2A, and shared state work together to connect multi-framework agents—and why Redis is the context layer that keeps them coherent.
Vector embeddings explained: from theory to real-world use
Learn how vector embeddings work, how similarity search scales, and where they power semantic search, RAG, and AI agents in production systems.
Dynamic chunking for RAG: building context infrastructure that adapts
Learn why fixed-size chunking caps RAG retrieval quality, how adaptive strategies fit into your context infrastructure, and how Redis helps you optimize retrieval in real time.
Best databases for agent memory: Redis vs. Pinecone vs. MongoDB vs. Weaviate
AI agents need three kinds of memory. Compare Redis, Pinecone, MongoDB and Weaviate on session state, long-term retrieval and operational state.
Multi-step AI agents: what they are & how they work
Learn how multi-step AI agents use the plan-act-observe loop, why they fail in production, and what your data layer needs to keep them reliable.
What’s new in two: June 2026 edition
Welcome to “What’s new in two,” your quick hit of Redis releases you might have missed in the past month.
Does quantization speed up inference?
Quantization can speed up LLM inference—but results vary by hardware, format, and batch size. Learn what works, what it costs, and how to stack it with caching.
Cache layer architecture: a practical guide to speed & scale
Learn how cache layers work, where they sit in your architecture, which caching patterns to use, and how to scale across nodes and regions without outages.
Tail latency: why the slowest requests matter most
Averages hide slow requests hurting your users. Learn what tail latency is, why fan-out amplifies it, and how in-memory architecture tightens your p99.
Agentic AI testing guide: methods & best practices
Learn how to test AI agents beyond model outputs—covering trajectory evaluation, tool testing, observability, and infrastructure for production reliability.
Semantic overload: why AI agents get facts wrong
Semantic overload degrades AI agent accuracy even with the right context. Learn why vector search alone falls short and what approaches close the relational gap.
Build AI agents with short-term & long-term memory in Redis
Learn how to build AI agents with short-term and long-term memory using Redis and LangGraph, from checkpointing to vector-based retrieval.
Comparing the best open source vector databases
See how Milvus, Weaviate, Qdrant, Chroma, pgvector, and Faiss compare on indexing, hybrid search, and licensing, and where a unified platform like Redis fits.
Token efficiency: getting more signal into the context window
Learn why more tokens hurt LLM reasoning, where low-signal noise comes from, and how reranking, hybrid search, and semantic caching improve output quality.
LLM router architecture: best practices for 2026
Learn how LLM model routers work, the three routing strategies that hold up in production, and how a real-time data layer fits into your router stack.
Knowledge graph retrieval-augmented generation (RAG): structured retrieval for AI agents
Learn how knowledge graph RAG solves multi-hop retrieval, graph freshness, and context rot—and where structured retrieval fits in an AI agent's memory stack.
Context engineering vs prompt engineering: the real difference
Prompt tweaks can't fix stale data or broken retrieval. Learn why context engineering—not prompt engineering—is the real lever for production AI agents.
Sub-agents: splitting context across specialized AI agents
Learn how sub-agents solve context window limits, why coordination fails without shared memory, and how Redis keeps multi-agent systems coherent at scale.
Dynamic batching: a practical how-to guide
Learn how dynamic batching improves GPU utilization, how it differs from continuous batching for LLMs, and how semantic caching cuts requests.
FAQ: Real-time context engine, agent memory, and retrieval
How to give AI agents the right context: RAG vs. agentic RAG, MCP, agent memory, and semantic caching explained. A practical FAQ for engineering teams.
Retrieval vs. memory in AI agents: why context layers need both
Retrieval answers what's in your data. Memory answers what happened before. Production agents need both—here's why stitching them together breaks.
Why a bigger context window won't fix your agent's memory
Larger context windows delay token limits but don't create persistent memory. Learn why agent memory needs its own persistence layer—and how to build it.
Your agent hit the context limit. Here's the playbook
Agent hitting its context limit? This six-step playbook trims tool outputs, compacts history, and moves durable state to fast external storage.
Five context-engineering principles that survive production
Context bloat, stale history, and bad retrieval break production agents. These five principles help you build AI agents that stay reliable at scale.
What Model Context Protocol gives your agents & what it leaves to you
MCP standardizes tool calls but not memory, freshness, or reliability. Learn what the protocol leaves to you and where a data layer fills the gap.
AI agent access control: a practical guide
Learn how to scope AI agent permissions per action, filter retrieval pipelines, and enforce field-level access controls to prevent data leaks in AI systems.
Connect Your Redis index to AI agents with RedisVL MCP
Developers love Redis. Unlock the full potential of the Redis database with Redis Enterprise and start building blazing fast apps.
Connecting to Redis Cloud with AWS PrivateLink vs. VPC peering
Developers love Redis. Unlock the full potential of the Redis database with Redis Enterprise and start building blazing fast apps.
Context windows in AI: why every token is a budget decision
Context windows are a budget, not a fill limit. Learn how longer contexts raise costs and hurt reasoning—and how to keep your AI context lean.
Real-time fraud detection for financial transactions
Fraud detection is a latency race you don't control. Learn how feature stores, sliding windows, and Redis keep scoring fast at billions of events.
Context engineering for AI agents: the infrastructure behind every decision
Context engineering decides what your agent can see at decision time. Learn the six inputs competing for every token budget and why fast assembly matters.
Redis Data Integration in Redis Cloud is now GA in AWS
Developers love Redis. Unlock the full potential of the Redis database with Redis Enterprise and start building blazing fast apps.
AI context windows: Why context quality beats context size
Poor context quality causes most AI agent failures, not model issues. Learn how retrieval precision, freshness, and hybrid search keep your LLM on track.
Redis vs Memorystore: key differences in 2026
Redis 8 and managed Google Cloud services have diverged since 2024. Compare features, vector search, and portability to find the right fit.
Semantic layer vs context layer: where BI modeling ends & AI grounding begins
Semantic layers standardize BI metrics. Context layers ground AI agents at runtime. Learn why reliable agents need both—and where Redis fits in.
AI reasoning explained: smarter models still need context
AI reasoning models think harder, not smarter. See the five production failure modes and why context quality determines output—not model intelligence.
Announcing Redis 8.8: New array data structure, rate limiter, performance improvements, & more
Redis 8.8 is now available in Open Source. Explore the new array data type, window counter rate limiter, streams NACK, and more performance-focused updates.
Redis 8.8 performance improvements: Faster MGET, MSET, streams, & more
Redis 8.8 is now available in Open Source. Explore the new performance improvements around a host of commands, persistence, and replication.
Diving deep into Redis’s new array data type
Redis 8.8 introduces a new array data type with O(1) index access. Learn how it works under the hood, what you can build with it, and when to use it over other types.
Why your AI doesn't understand your business (& how teams fix it)
AI fails in production when business context is stale, noisy, or conflicting. Learn how teams fix context gaps with fast retrieval and real-time data.
Context graphs: when nearest-neighbor search isn't enough
Context graphs traverse entity relationships ANN search can't follow. Learn when to use graphs, vectors, or both in your RAG retrieval stack.
What’s new in two: May 2026 edition
Welcome to “What’s new in two,” your quick hit of Redis releases you might have missed in the past month.
How Conflict-free Replicated Data Types power active-active database replication
Learn how Conflict-free Replicated Data Types solve write conflicts in active-active replication and why Redis uses CRDTs for Geo Distribution.
Single-shot reliable consumers with XREADGROUP CLAIM in Redis 8.4
Developers love Redis. Unlock the full potential of the Redis database with Redis Enterprise and start building blazing fast apps.
Context orchestration: what it is & how it works
Learn what context orchestration is, how it differs from RAG and prompt engineering, and how Redis powers the context engine layer for production AI systems.
Context compaction for AI agents: a complete guide
Learn how context compaction keeps AI agents on task as sessions grow—and how Redis supports memory, caching, and retrieval in one platform.
Prompt bloat: causes, costs & fixes for LLM apps
Prompt bloat slows LLM apps, raises costs, and hurts quality. Learn the root causes and how context engineering with Redis can fix it.
Agentic retrieval techniques: a complete guide
Learn key agentic retrieval techniques—hybrid search, routing, query planning, and caching—and how Redis powers the context layer beneath them.
Long-horizon tasks: building agents that work over hours & days
Learn why long-horizon agents fail and how durable memory, checkpointing, and Redis Agent Memory keep agents running across hours and days.
What is a context engine? The platform layer behind production AI agents
Learn what a context engine is, how it fits into agent architecture, where RAG falls short, and how Redis powers the context layer in production.
What is a context layer? AI agent infrastructure
Learn what a context layer is, how it prevents agent failures, and how Redis Iris delivers managed context infrastructure for reliable AI agents.
Context is all you need: Introducing Redis Iris
Developers love Redis. Unlock the full potential of the Redis database with Redis Enterprise and start building blazing fast apps.
Context retrieval for AI agents: what it is & why it matters
Learn how context retrieval works in AI agents, why basic RAG fails at scale, and how Redis supports reliable retrieval with hybrid search and memory.
Context poisoning: how bad information breaks agent reasoning
Context poisoning corrupts AI agent reasoning silently. Learn how it spreads through RAG, memory, and tools—and how to keep agent context fresh.
How to tame the thundering herd problem
The thundering herd problem occurs when multiple processes or clients repeatedly request the same resource simultaneously, leading to excessive load
Context engineering for AI: what it is & how to build it
Context engineering is the discipline of managing everything an LLM receives during inference. Learn what it is, why it matters, and how to build it.
Dynamic endpoints: Migrate databases without changing your endpoint
Developers love Redis. Unlock the full potential of the Redis database with Redis Enterprise and start building blazing fast apps.
AI shopping assistants: how they work & what to build
Learn what AI shopping assistants are, the five types, and the infrastructure needed to build one that's fast, fresh, and conversion-ready.
Endless aisle retail: infrastructure & real-time data
Endless aisle retail connects stores to full catalogs in real time. Learn the infrastructure challenges, AI trends, and pitfalls to avoid when building one.