Getting Started with Mem0 + Valkey

What is Mem0?

Mem0 ("mem-zero") is an open-source memory layer for AI applications (YC S24). It has a dedicated Valkey connector (provider: "valkey") that uses the valkey Python client with native FT.CREATE/FT.SEARCH for vector-based memory storage and retrieval.

Mem0 has a first-class Valkey integration - not just Redis compatibility. The valkey.py connector in Mem0's source uses the native valkey Python package and supports HNSW/FLAT indexing with configurable parameters.

Step 1: Install

pip install mem0ai

The valkey Python client is included as a core dependency - no separate install needed.

Step 2: Start Valkey with Search Module

# Valkey with valkey-search module loaded
valkey-server --loadmodule /usr/lib/valkey/libsearch.so

Step 3: Configure Mem0 with Valkey

from mem0 import Memory

config = {
    "vector_store": {
        "provider": "valkey",
        "config": {
            "valkey_url": "valkey://localhost:6379",
            "collection_name": "mem0",
            "embedding_model_dims": 1536,
            "index_type": "hnsw",
        }
    }
}

memory = Memory.from_config(config)
print("Mem0 connected to Valkey!")

Step 4: Add Memories

# Add memories from a conversation
messages = [
    {"role": "user", "content": "I love Italian food, especially pasta carbonara."},
    {"role": "assistant", "content": "Great choice! Carbonara is a classic Roman dish."},
]
result = memory.add(messages, user_id="user_001")
print(f"Added: {result}")

# Add more context
messages2 = [
    {"role": "user", "content": "I'm allergic to shellfish and prefer spicy food."},
    {"role": "assistant", "content": "Noted! I'll keep that in mind for recommendations."},
]
memory.add(messages2, user_id="user_001")

Step 5: Search Memories

# Search for relevant memories
results = memory.search(
    query="What food does this user like?",
    user_id="user_001",
    limit=3,
)

for entry in results["results"]:
    print(f"Memory: {entry['memory']}")
    print(f"Score: {entry.get('score', 'N/A')}\n")

# Output:
# Memory: Loves Italian food, especially pasta carbonara
# Score: 0.87
# Memory: Allergic to shellfish, prefers spicy food
# Score: 0.62

Step 6: Get All Memories for a User

# Retrieve all memories for a user
all_memories = memory.get_all(user_id="user_001")
for m in all_memories["results"]:
    print(f"  - {m['memory']}")

Step 7: Use Memories in a Chatbot

from openai import OpenAI

openai_client = OpenAI()

def chat_with_memory(message: str, user_id: str) -> str:
    # 1. Retrieve relevant memories
    relevant = memory.search(query=message, user_id=user_id, limit=3)
    memories_str = "\n".join(
        f"- {entry['memory']}" for entry in relevant["results"]
    )

    # 2. Build prompt with memories
    system = f"You are a helpful AI. Use these memories about the user:\n{memories_str}"
    messages = [
        {"role": "system", "content": system},
        {"role": "user", "content": message},
    ]

    # 3. Generate response
    response = openai_client.chat.completions.create(
        model="gpt-4", messages=messages,
    )
    answer = response.choices[0].message.content

    # 4. Save new memories from this conversation
    memory.add(
        [{"role": "user", "content": message},
         {"role": "assistant", "content": answer}],
        user_id=user_id,
    )
    return answer

# Usage
response = chat_with_memory("Recommend me a restaurant", "user_001")
print(response)
# Will recommend Italian restaurants, avoid shellfish, suggest spicy options!

How It Works Under the Hood

Source: mem0/vector_stores/valkey.py - The official Valkey connector in the Mem0 repository.

Next →02 - Multi-User Memory