Events, embedded as they land.
Pick a text column and an Nimbia embedding model. Basalt embeds new rows on arrival and stores each vector next to its source row.
How it works.
Basalt batches new rows and calls the Nimbia Embeddings API with your key. Vectors are written to a vector column on the same row, and each row is embedded once, keyed by a hash of its text.
Set it up in three steps.
Connect the tool
Store your Nimbia API key as a workspace secret and choose the view and column to embed. Backfill runs at your rate limit.
# embed support ticket bodies with text-embedding-3-small basalt connect openai --view tickets --column body --model text-embedding-3-small ✓ vector column body_embedding added · 1536 dims ✓ backfill 284k rows · 3m 12s
Shape it with a view
Write plain SQL over the source. Basalt keeps the view current with every record and tracks lineage back to the topic, partition and offset.
CREATE MATERIALIZED VIEW revenue_by_min AS SELECT date_trunc('minute', ts) AS minute, currency, sum(amount) AS revenue, count(*) AS payments FROM payments WHERE status = 'captured' GROUP BY 1, 2;
Read it from anywhere
Every view is an endpoint the moment it exists. Call it over REST, connect with any Postgres client, or use the SDK from your app.
curl https://api.basalt.dev/v1/views/revenue_by_min?limit=5 \ -H "Authorization: Bearer $BASALT_TOKEN"
What you get.
Embed once
Rows are keyed by content hash, so unchanged text is never re-embedded.
Vectors beside rows
Search by similarity and filter by any column in the same query.
Rate aware
Requests respect your tier limits and back off on 429 responses.
Model choice
Use text-embedding-3-small or -large, with optional reduced dimensions.
Set your data in stone.
Free up to 10M events a month. No card, no sales call. Your first pipeline is live in about four minutes.