> ## Documentation Index
> Fetch the complete documentation index at: https://docs.cf0.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Data governance

> Where your firm's data lives in cf0, what we store and don't, no-training guarantees, org isolation, data residency, and disclosed sub-processors.

cf0 is built so analysts can put their name on the output. That starts with knowing exactly where data lives and who can touch it.

## Storage

| What                                 | Where                                              |
| ------------------------------------ | -------------------------------------------------- |
| Threads, messages, generated reports | Immutable per-org storage, US region               |
| User accounts, org membership, roles | Enterprise identity layer with a relational record |
| Uploaded documents                   | Org-scoped object storage, US region               |
| SEC + global filings index           | Structured, queryable index                        |

Threads and reports live as an immutable canonical record, not in a mutable database or codebase. Deletion is immediate from the user's view; backing storage is purged on the next sync cycle.

## Sub-processors

cf0 uses a curated set of sub-processors covering:

* **Cloud infrastructure** — primary compute, storage, and AI inference, US region
* **Identity and access** — enterprise authentication and organisation management
* **Durable workflow orchestration** — long-running ingestion and processing
* **Frontend hosting** — global edge delivery
* **Product analytics** — usage telemetry (no thread or report content)
* **AI observability** — model I/O tracing (content stays inside the platform perimeter)
* **Semantic retrieval** — vector index for source lookup
* **Data feeds** — read-only ingest of filings, market data, transcripts, and macro data; do not process customer data

The complete sub-processor list — vendor, role, and data scope per processor — is disclosed under the DPA. Contact [filippo@cf0.ai](mailto:filippo@cf0.ai) to request it.

## Boundaries

* **Org isolation** — every thread, document, and report is scoped to your organisation. There is no cross-org visibility.
* **No training on your data** — cf0 never uses customer threads, documents, or reports to train any AI model.
* **Inference inside the platform perimeter** — AI inference runs inside cf0's cloud infrastructure; customer data does not leave the platform to reach a model.

## Compliance

For regulated workflows, the DPA, or sub-processor disclosure, contact [filippo@cf0.ai](mailto:filippo@cf0.ai).