Upsert

Write documents to both stores in one call.

Each document carries an id, either text (which gets embedded) or a vector, and optional metadata. dynavec splits metadata: a small filterable subset goes to S3 Vectors, the full copy plus text goes to DynamoDB.

from dynavec import Document

db.upsert([
    Document(id="1", text="apple pie recipe", metadata={"cat": "food", "rating": 5}),
    Document(id="2", text="rocket launch schedule", metadata={"cat": "space"}),
], namespace="kb", auto_metadata=True)

The metadata switch

You provide metadata — stored verbatim.
auto_metadata=True — dynavec also derives created_at, content_hash, word_count, char_count. Your keys win on conflict.

Re-upserting the same id overwrites it. Writes to the two stores run in parallel; see Concurrency. To change part of a document, use update.

Document size limit

Each document's text and metadata are stored as one DynamoDB item, which DynamoDB caps at 400 KB. upsert() and update() check every document's size before writing anything, and raise ItemTooLargeError (with doc_id, namespace, size_bytes, and limit_bytes) if one is too big. The whole call fails, so S3 Vectors and DynamoDB never end up with half a batch.

from dynavec import ItemTooLargeError
from dynavec.ingest import chunk_text

try:
    db.upsert([Document(id="manual", text=long_text)], namespace="kb")
except ItemTooLargeError as err:
    print(err.doc_id, err.size_bytes, err.limit_bytes)
    db.upsert(
        [Document(id=f"manual#chunk{i}", text=chunk)
         for i, chunk in enumerate(chunk_text(long_text, chunk_size=2000))],
        namespace="kb",
    )

PreviousEmbeddings NextUpdate & Lambda