How does Google Vertex AI Search improve internal knowledge discovery?

All answers to this question.

Vertex AI Search is designed exactly for this "silo-busting" scenario. It provides "out-of-the-box" connectors for Google Workspace, Salesforce, and Jira, making the initial indexing process very straightforward. The most impressive feature is "ACL (Access Control List) synchronization." It inherits the permissions from the source system; if an employee doesn't have access to a specific Doc in Google Drive, that Doc will not appear in their search results, nor will the AI use it to generate answers for them. This solves the major security headache of internal AI search.

   Answered 2024-01-22 by Helen Robinson

Have you considered how you'll handle "stale" data? If you index everything, the AI might give an answer based on an outdated policy from 2019 instead of the 2024 version.

   Answered 2024-01-23 by Steven Hall

• Steven, that’s a valid point. We plan to use the "Data Store" filtering feature in Vertex AI Search to only index folders marked as "Current Version." Helen, the ACL sync is the deal-breaker for us. We were worried about payroll info leaking into general searches. Knowing that the AI respects the existing Drive permissions means we can deploy this much faster without a complete overhaul of our folder structures. This is going to save our engineering team at least 5 hours of 'search time' a week.

     Commented 2024-01-24 by Elizabeth Taylor

You can also embed the search widget directly into your internal company portal with just a few lines of code. It’s very "plug-and-play" compared to building a custom RAG app.

   Answered 2024-01-25 by Nancy Young

• Nancy is right. The low-code nature of Vertex AI Search means our HR team can actually manage the search results without constantly bothering the IT department.

     Commented 2024-01-26 by Helen Robinson