> ## Documentation Index
> Fetch the complete documentation index at: https://docs.negentro.tech/llms.txt
> Use this file to discover all available pages before exploring further.

# Manage information throughout its lifecycle

> Reliable AI memory depends on more than storing records. Your application needs a consistent approach to data quality, updates, access, retention, deletion, and recovery.

## The data lifecycle

## Organize data with metadata

| Field | Purpose |
| :- | :- |
| type | Record type or category for classification and filtering. |
| source | Source identifier for provenance tracking. |
| created\_at / updated\_at | Timestamps for freshness evaluation. |
| user\_id / session\_id | User, session, or tenant scope association. |
| project\_id | Project or workflow identifier for contextual grouping. |
| retention\_category | Classification for applying retention and deletion rules. |
| version / status | Version or lifecycle status tracking. |

## Data quality and consistency

Define how your application handles duplicate records, conflicting facts, outdated values, and updates from multiple sources. Where information changes over time, preserve enough provenance to explain where a fact came from and when it was observed.

## Retention and deletion

Create clear rules for how long different types of information should remain available. A deletion workflow should account for the authoritative record, derived summaries, search indexes, caches, and other copies where applicable.

## Backups and recovery

Document backup frequency, retention periods, restore procedures, recovery objectives, and the responsibilities of your team versus your infrastructure provider.

## Operational checklist


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.