Vertex AI Agent Builder Advanced Document Data Model

Google Cloud Build with AI Custom ML & MLOps
1

We have some questions about the Vertex AI Agent Builder Advanced Document Data Model.

https://cloud.google.com/generative-ai-app-builder/docs/configure-autocomplete?hl=en&_gl=111vv0bj_gaMjA0MDQ3MDk1NS4xNzMwMjEzMDAw_ga_WH2QY8WWF5*MTczNTMyNjE4Ny42OS4xLjE3MzUzMjg0OTUuNTguMC4w#advanced-document

  1. What’s the process to enable and set up the advanced document model for web-crawled sites? Do we need to update the schema or re-crawl the website to make it work?
  2. If we turn on the advanced model, do we have to make any changes to how data is imported or to the schema, or does it just work with the data we’ve already got?
  3. What kind of trade-offs should we expect in terms of speed or performance when using the advanced document model for autocomplete compared to the regular document or completable fields model?
  4. If our website has a mix of structured and unstructured content, how well does the advanced model handle prioritizing the most relevant data for autocomplete suggestions?
  5. Is it worth enabling the advanced model for a web-crawled website, or would the standard models be sufficient in most cases?
2

Hi @imgali,

Welcome to Google Cloud Community!

Based on the documentation:

This feature is experimental. If you’re interested in using this feature, contact your Google Cloud account team and ask to be added to the allowlist.

Was this helpful? If so, please accept this answer as “Solution”. If you need additional assistance, reply here within 2 business days and I’ll be happy to help.