Adds a new vector and full-text search quickstart - #7923
Conversation
Elastic Docs AI PR menuCheck the box to run an AI review for this pull request.
Powered by GitHub Agentic Workflows and docs-actions. For more information, reach out to the docs team. |
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
🔍 Preview links for changed docs |
Elastic Docs Style Checker (Vale)Summary: 2 warnings found
|
| File | Line | Rule | Message |
|---|---|---|---|
| solutions/vector-database/vector-full-text-search.md | 512 | Elastic.MenuArrows | Use '→' to separate menu items, not '' or '='. Example: Select Manage index → Add lifecycle policy. |
| solutions/vector-database/vector-full-text-search.md | 512 | Elastic.MenuArrows | Use '→' to separate menu items, not '' or '='. Example: Select Manage index → Add lifecycle policy. |
The Vale linter checks documentation changes against the Elastic Docs style guide. To use Vale locally or report issues, refer to Elastic style guide for Vale.
|
Hi @alexmarhaba @dustincoates @georgewallace @KOTungseth @leemthompo , I've added the tutorial here, so please feel free to review it from now on. These commits include the modifications we agreed on in our last call. All commits include additional content refinements. I ran our review guidelines against this content and found a few areas that I thought would benefit from some small tweaks based on our guidelines. These include:
All these changes are negotiable. Please let me know what you think. Thank you in advance! |
| ::::{tab-item} macOS and Linux | ||
|
|
||
| ```bash | ||
| export ES_URL="https://YOUR-PROJECT.es.REGION.aws.elastic.cloud:443" |
There was a problem hiding this comment.
We need to tell folks how to find this: https://www.elastic.co/docs/solutions/elasticsearch-solution-project/search-connection-details
(#7709 is updating that page)
There was a problem hiding this comment.
In this step we advise users to copy those from the Getting started page they land on when they create a new project. I was also considering adding guidance for cases where they navigate to other pages, but I was afraid that would make the text unnecessarily longer. I'm leaning towards what we have in the Create a project step being enough, but I'm happy to add the Search connection details page if you'd suggest it.
|
I took a quick glance and it looks good to me, thanks! |
| navigation_title: Vector and full-text search | ||
| description: Step-by-step tutorial for connecting an application to an Elasticsearch Vector Database project, indexing data, and running full-text, semantic, hybrid, and ES|QL searches. | ||
| applies_to: | ||
| serverless: ga |
There was a problem hiding this comment.
If this page is intended for this particular project type, then we can use the new product id I've created for it.
See more details in https://github.com/elastic/docs-content-internal/issues/1534
There was a problem hiding this comment.
Also, this is now merged and should be released soon, so with the new version of docs-builder, you'll be able to tag the page with the applies_to tag as well.
|
After testing the tutorial end to end thoroughly, I found that the previous structure mixed two usage models. Users were expected to add snippets to one file and rerun it for verification, but rerunning attempted to recreate the books index and failed with resource_already_exists_exception. Individual examples also depended on earlier imports and initialization, so they could not be tested independently. The guide was still usable when followed as a single uninterrupted flow, but the overall user experience did not feel clear or intuitive. Since we agreed that each step should be independently verifiable to support users’ understanding and learning, I updated the tutorial to provide two clear paths:
Happy to tweak anything based on feedback. |
This PR adds a new vector and full-text search quickstart that:
Testing
All seven client examples were tested end to end with Claude against Elasticsearch. The tests verified the expected search rankings and ES|QL aggregation output.
Manually testing the tutorial: WIP
Generative AI disclosure
Claude, Cursor - GPT 5.6