Memory Brain

How Lead360 helps

Memory Brain gives a sub-account a governed knowledge layer for reviewed business context. Every change creates a revision with normalized content and a checksum, reviewers approve exact revisions, and available snapshots preserve the approved state used to build immutable Memory Sets.

When to use this area

Use Memory Brain for stable business knowledge that a Piece or Ask360 bot should reference, such as policies, service explanations, operating procedures, and reviewed website content.

Before you begin

• Remove passwords, API keys, payment details, and other credentials from source files.
• Decide which knowledge is stable enough for reviewed memory and which values must still be queried live.
• Identify the reviewer responsible for the exact sub-account and subject area.

What this area does

Memory Brain stores reviewed business context for this exact sub-account. It does not grant tools, permissions, tenants, profiles, or write targets.

A reliable workflow

1. Create or update a pending YAML revision.
2. Review the normalized content, evidence, and checksum.
3. Approve the exact revisions and bind one immutable Memory Set to the intended Piece.

Confirm the result

A completed Memory Brain workflow should be visible in the product area that owns the final record. Use these checks before relying on the result or moving to the next customer-facing step.

• Review normalized content, source evidence, revision state, and checksum.
• Make the intended reviewed revisions available as a new snapshot.
• Bind only the required files through a Memory Set or Learning Collection and test the consumer.

Troubleshooting

A file stays pending

A pending revision has not completed the review and availability workflow. Mark the exact revision reviewed, then create the next available snapshot.

A live value is outdated

Prices, availability, booking links, payment status, and other changing values should come from live Lead360 tools rather than static memory files.