AI-200
Azure
Back to study
Data Management for AI(25–30%)

Azure Managed Redis for AI Caching

Cache embeddings, session data, and vector indexes with Azure Managed Redis.

Overview

Azure Managed Redis provides low-latency caching and can support vector similarity search for AI retrieval workloads.

Key concepts

  • Caching patterns — cache-aside, write-through for embedding results
  • TTL and eviction — volatile-lru for expiring cache keys
  • Vector indexing — Redis vector search capabilities for similarity queries
  • Connection resilience — retry policies, connection multiplexing

Exam tips

  • Know when Redis complements (not replaces) Cosmos DB / PostgreSQL
  • Understand cache invalidation strategies for updated documents
  • Vector index configuration for dimensions and distance metric

Azure CLI

Azure CLI
# Create Azure Managed Redis (preview name may vary by region)
az redis create \
  --name redis-ai200 \
  --resource-group rg-ai200 \
  --location eastus \
  --sku Basic \
  --vm-size c0

# Get connection string
az redis list-keys \
  --name redis-ai200 \
  --resource-group rg-ai200

Python — cache embeddings

Python
import redis
import json

r = redis.Redis(
    host="redis-ai200.redis.cache.windows.net",
    port=6380,
    password=access_key,
    ssl=True,
)

def get_embedding_cached(text: str) -> list[float]:
    key = f"emb:{hash(text)}"
    cached = r.get(key)
    if cached:
        return json.loads(cached)

    embedding = call_embedding_model(text)  # Azure OpenAI, etc.
    r.setex(key, 3600, json.dumps(embedding))
    return embedding

Learn more