Skip to main content
Wrap memory backends with Redis caching for improved performance on frequently accessed conversations.

​
Installation

Install the required packages: 
pip install hypertic[redis]

​
Setup

Set your Redis connection string: 
export REDIS_URL=redis://localhost:6379

​
Usage

Redis Cache wraps other memory backends to provide caching. It works with both PostgreSQL and MongoDB, and supports both sync and async operations: 
import os
from dotenv import load_dotenv
from hypertic.agents import Agent
from hypertic.models import OpenAIChat
from hypertic.memory import PostgresServer, RedisCache

load_dotenv()

DB_URL = os.getenv("DATABASE_URL", "postgresql://postgres:password@localhost:5432/postgres")
REDIS_URL = os.getenv("REDIS_URL", "redis://localhost:6379")

def main():
    # Step 1: Create base memory backend (PostgreSQL)
    postgres_memory = PostgresServer(db_url=DB_URL, table_name="cache_memory")

    # Step 2: Wrap with Redis cache
    # TTL: 1 hour (3600 seconds)
    cached_memory = RedisCache(
        store=postgres_memory,
        redis_url=REDIS_URL,
        ttl=3600,
        key_prefix="trial"
    )

    # Step 3: Use with Agent
    agent = Agent(
        model=OpenAIChat(
            model="gpt-4o-mini"
        ),
        memory=cached_memory
    )

    user_id = "user_123"
    session_id = "session_001"

    # Non-streaming response
    response = agent.run("My favorite color is blue.", session_id=session_id, user_id=user_id)
    print(response.content)

    # Streaming response
    for event in agent.stream("what is my favorite colour?", session_id=session_id, user_id=user_id):
        if event.type == "content":
            print(event.content, end="", flush=True)

if __name__ == "__main__":
    main()

​
Configuration Options

  • store - The base memory backend to wrap (PostgresServer, MongoServer, or their async equivalents)
  • redis_url - Redis connection string
  • ttl - Time-to-live for cached entries in seconds (default: 3600)
  • key_prefix - Prefix for Redis keys (default: “agent”)
Redis Cache can wrap any memory backend (PostgreSQL or MongoDB) to provide a caching layer. For async operations, use AsyncRedisCache with AsyncPostgresServer or AsyncMongoServer. You can inspect cached data using Redis CLI commands like KEYS trial:* or GET trial:key.
For more details, see the Redis documentation.