Should I use remember or document ingest for this?

I have both structured facts and raw chat logs to store. Which method should I use for each?

Unison has two paths for getting content into the brain. They serve different purposes and have different latency characteristics.

Conversation capture: POST /v1/brain/remember (async curation)

POST /v1/brain/remember
# body: {"dump":{"turns":[{"role":"...","content":"..."}]},"source":"...","sourceRef":"..."}

dump can also be a raw string, or {"sessionJsonl": "..."} for a raw session log.

Best for: raw agent-user exchanges, support transcripts, meeting notes — any unstructured conversational text where you want Unison to figure out what's worth keeping.

• Returns { "jobId": "..." }. Processing is async (a save-or-skip curation pass: dedupes against existing memory, then files curated notes under /private/notes/ and extracts entity facts).
• Poll GET /v1/brain/jobs/{id} for status.
• Recall over the curated notes/facts becomes available once the job completes.
• More powerful, but you don't control exactly what gets kept or how it's filed.

Document ingest / write (synchronous)

POST /v1/brain/ingest
# body: {"items":[{"type":"document","content":"...","title":"...","path":"/private/notes/my-doc.md"}]}

PUT /v1/brain/doc
# body: {"path":"/private/notes/my-doc.md","bodyMd":"...","title":"..."}

Best for: structured knowledge you've already curated — decisions, policies, FAQs, reference data.

• Available for recall immediately after the write returns (ingest for documents is synchronous; PUT /v1/brain/doc is a single-document convenience over the same write path).
• You control the content exactly — no extraction, no curation pass.
• Useful when you need deterministic, precise information in the brain.

Rule of thumb

If you're feeding raw conversational data and want Unison to decide what's worth remembering, use remember. If you already know what the fact is and want it in the brain right now, ingest or write it as a document.

Many applications use both: remember live agent conversations automatically, and write hand-authored knowledge pages for things like team decisions or onboarding context.