Skip to main content

​
Operational Pipeline

​
The ingestion API operates on an asynchronous, event-driven model.\

Requests are validated and acknowledged immediately (202 Accepted), then offloaded to a background worker queue. The data undergoes a specific transformation sequence before becoming searchable:
  1. Shift-Left Temporal Resolution: The pipeline scans source.content for relative time references (e.g., “next Friday”, “in three days”). These are resolved into absolute ISO 8601 timestamps at the moment of ingestion, creating a deterministic temporal index.
  2. Semantic Enrichment: The system applies bidirectional expansion, injecting implicit context (themes, synonyms) to maximize vector overlap.
  3. Dual-Write Persistence: Data is synchronously committed to both the Metadata Store and Vector Database.
  4. Centroid Calibration: Finally, the user’s semantic centroid is recalculated to reflect the new memory distribution, influencing future adaptive search modes.
Memory ingestion pipeline

​
Text ingest (API)

Submit a text-based memory payload.

​
Payload schema (text)

FieldTypeStatusDescription
source.typeStringREQUIREDMust be text.
source.contentStringREQUIREDThe raw text to ingest. This field is parsed for automatic relative date resolution (Shift-Left).
userContextStringOPTIONALAdditional meta-context (e.g., “Slack Message”, “Email from CEO”) used to guide the enrichment engine.
curl -X POST "https://api.memorymodel.dev/v1/ingest" \
  -H "Content-Type: application/json" \
  -H "x-memory-cluster-access-key: sk_live_..." \
  -H "x-end-user-id: user_123" \
  -d '{
    "source": {
      "type": "text",
      "content": "The project deadline has been moved to next Friday."
    },
    "userContext": "Slack message from Project Manager"
  }'
Include x-end-user-id to track the end-user identity. This header is required in multi‑user setups; if your cluster is single‑user, you can use a simple identifier like user_123.
Response (text) 
{
  "status": "Accepted",
  "jobId": "job_1700000000000_user_123"
}

​
Image ingest (API)

Submit an image for multimodal processing. The system extracts visual features and semantic context, integrating them into the memory graph alongside text nodes.

​
Payload schema (image)

FieldTypeStatusDescription
imageDataStringREQUIREDBase64-encoded image string.
userContextStringOPTIONALContext to anchor the visual analysis (e.g., “Screenshot of error logs”).
curl -X POST "https://api.memorymodel.dev/v1/ingest/image"  \
  -H "Content-Type: application/json" \
  -H "x-memory-cluster-access-key: sk_live_..." \
  -H "x-end-user-id: user_123" \
  -d '{
    "imageData": "iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAQAAAC1HAwCAAAAC0lEQVR42mNk+A8AAQUBAScY42YAAAAASUVORK5CYII=",
    "userContext": "Screenshot of the error message"
  }'
For large images, ensure your client supports large payloads. The imageData field must be a valid Base64 string (with or without the data URI prefix).
Response (image) 
{
  "status": "Accepted",
  "jobId": "img_job_1700000000000_user_123",
  "previewUrl": "https://storage.googleapis.com/..."
}
The previewUrl must point to a valid storage bucket and object. Invalid bucket names will return an error from the storage provider.