Log inSign up
Redis profile banner
user avatar

Redis

@Redisinc
See how fast feels
redis.io
Joined January 2014
2,994
Following
44K
Followers
RepliesMedia
  • user avatar
    Redis
    @Redisinc
    17h
    Most teams debug a wrong RAG answer by rewriting the prompt. That's usually the wrong place to look. The real culprit is often retrieval, ranking, or freshness, not the model. Redis Iris keeps all three on one real-time platform.
    Image
  • user avatar
    Redis
    @Redisinc
    Aug 10
    MCP vs. A2A isn't a headcount question. MCP means an agent calls down to a passive tool and waits for an answer. A2A means independent agents, often owned by different teams, delegate work to each other without exposing their internal logic. 🧵
    Image
  • user avatar
    Redis
    @Redisinc
    Aug 7
    The best AI apps run on better context, not better prompts. Every request assembles system instructions, history, retrieved knowledge, memory, and tool outputs into one prompt. The hard part is assembling the right context. That's why we build Redis Iris: redis.io/blog/context-a…
    Image
  • user avatar
    Redis
    @Redisinc
    Aug 6
    Your AI agent forgot everything you told it. Again. Redis Agent Memory just went live on the Snowflake Marketplace. Cortex agents are strong and can reason well, but ask one to recall what a user said last week, and it draws a blank. Every session starts from zero. 🧵
    Image
  • user avatar
    Redis
    @Redisinc
    Aug 5
    Keyword-based routing only gets you so far. When your system needs to understand intent (not just match words), you need something smarter. When your routing logic lives in Redis, it updates in real time. Here's how to build it with Redis and Java: redis.io/blog/designing…
    Image