Quickstart

1. Install

pip install mpalo

2. Authenticate

Get your API key from the Mind Platform → API Keys.

from mpalo import Palo

client = Palo(api_key="your-api-key")

3. Basic Memory Workflow

Pass user input and a unique identifier for the user. Palo Bloom returns an enriched context block ready for your LLM.

result = client.mind(
    "What should I cook tonight?",
    user_id="user_123"
)

4. Advanced Retrieval

For more complex tasks, specify a retrieval mode. See Memory Engines for guidance on when to use each.

# Traversal -- episodic context across the user's timeline
result = client.mind(
    "What did we decide last week?",
    user_id="user_123",
    mode="traversal"
)

# Mapping -- multi-session synthesis at full depth
result = client.mind(
    "Synthesize my project progress.",
    user_id="user_123",
    mode="mapping"
)

5. Bulk Memory Ingest

Bootstrap user memory from documents, archives, exported conversations, transcripts, or other existing data before the first live request. Bulk Memory Ingest runs asynchronously with Encode on and Render, Traversal, Mapping, and High-Effort Dreaming off by default.

client.migrate(
    source="path/to/data.json",
    user_id="user_123",
    engine="mpalo-palo"
)

The SDK is actively developed. For the latest method signatures always refer to the Python SDK reference.

Next Steps