metabase / metabase/metabase

Generate vector embedding for filter widget inputs

Open
#73,768 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

.Auto triaged .Team/QueryingPlatform Querying/Parameters & Variables Type:New Feature
Dominant language
Clojure
Stars
49.3k
Forks
6.8k
Avg merge
1d 13h
Merged PRs (30d)
653

Description

**What problem will this feature request solve?**
You can't perform a semantic search of data you have connected to Metabase. You can only search connected data sources based on "contains". (Not this is different from the semantic search of Metabase content, which scans the descriptions of charts and dashboards, but not rows of connected data).

**Describe the solution you'd like.**
Add an option to generate a vector embedding when performing a contains search in Metabase, e.g. with a string filter in the dashboard. This would allow you to set up a SQL variable, pass in the embedding, and perform the search.

**How does this feature request impact you?**
This would be useful for looking up records in my database. For example, if I had a vectorized database of GitHub issues, I could search through those records.

**Additional information**
Some folks use Metabase as a data exploration tool. This could help their workflows

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by reviewing how dashboard string filter inputs handle "contains" searches and how SQL variables receive filter values. Compare that flow with Metabase's existing semantic search of charts and dashboards. Done means a user can opt to generate an embedding from a filter input and pass it into a SQL variable for searching connected data rows.

Written by the indexing model from the issue text.

Assessment

Tech stack
clojure
Domain
databases, search
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Quiet
Clarity
Needs clarification
Newbie friendliness
30/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.