How to use Agentspace search on Google Distributed Cloud

Agentspace Search on Google Distributed Cloud: Summary

Key Concepts:

• Agentspace Search: Google-quality search solution for on-premise environments.
• Google Distributed Cloud: Google's offering for bringing cloud services to on-premise infrastructure.
• Multi-modal Search: Search capabilities that understand and process various data types (text, images, etc.).
• Gemini: Google's AI model powering Agentspace Search.
• Data Connectors: Pre-built integrations to connect to various on-premise enterprise systems.
• ACL (Access Control List): Security mechanism to control user access to data.
• On-Premise: Refers to infrastructure and systems located within an organization's physical premises.

1. The Problem: Data Regulation and the Need for On-Prem Search

• Main Point: Enterprises face challenges due to restrictive data regulations and privacy concerns that prevent data from leaving their premises.
• Specific Detail: These enterprises still require robust search capabilities for their knowledge workers.
• Solution: Google Agentspace Search addresses this by bringing Google-quality search to on-premise environments.

2. Agentspace Search: Features and Functionality

• Main Point: Agentspace Search offers a solution for on-premise enterprises needing powerful search capabilities without violating data regulations.
• Key Features:
  • Google-Quality Search: Provides search results comparable to Google's online search.
  • Multi-modal: Powered by Gemini, it can understand and process different types of data.
  • Pre-built Data Connectors: Integrates with various on-premise enterprise systems (e.g., SharePoint, ServiceNow, Confluence) to break down data silos.
  • In-built ACL Controls: Ensures data security by restricting user access to authorized data only.

3. Administrator Setup: Data Connector Configuration

• Main Point: Administrators can easily configure data connectors to integrate Agentspace Search with various on-premise systems.
• Step-by-Step Process (SharePoint Example):
  1. Access the Google Distributed Cloud console.
  2. Select the desired data connector (e.g., SharePoint).
  3. Enter the required settings and configure advanced options.
  4. Choose specific data entities to sync and set the synchronization frequency.
• Flexibility: Administrators can set up multiple data connectors to ingest data from different systems.

4. End-User Experience: Financial Analyst Use Case

• Main Point: Agentspace Search provides an intuitive search UI for enterprise users to access and analyze data.
• Scenario: A financial analyst at a banking company needs to create a credit report.
• Step-by-Step Example:
  1. Initial Question: "When is this credit report due?" Agentspace Search retrieves the due date (April 15th) from relevant sources (emails, SharePoint, etc.).
  2. Follow-up Question: "Tell me what the top emerging credit risks are." Agentspace Search provides an AI overview, search results, and the data sources used to generate the information.
  3. Drilling Down: "Give me details on specific market volatility risks." Agentspace Search retrieves information from Confluence and other enterprise systems, formatting it neatly.
  4. Source Verification: Users can click on citations to see the specific documents used to generate the results, enhancing trust in the system.
  5. Client Impact: "Can you summarize a list of all client support tickets that were raised?" Agentspace Search pulls relevant tickets from ServiceNow.
• Outcome: The financial analyst has all the necessary information to build the credit report.

5. Gemini's Multi-Modal Capabilities

• Example: When detailing market volatility risks, Gemini formats the information neatly, showcasing its ability to process and present data effectively.
• Significance: This highlights the power of multi-modal AI in understanding and organizing complex information.

6. Importance of Source Citation

• Key Point: Agentspace Search provides citations to specific documents, allowing users to verify the source of information.
• Quote: "This is really important for users to be able to have trust in the system when you're able to point to very specific sources of on-prem information."
• Benefit: This builds trust and confidence in the accuracy and reliability of the search results.

7. Conclusion

• Main Takeaway: Agentspace Search on Google Distributed Cloud empowers on-premise enterprises with Google-quality search capabilities while adhering to data regulations and privacy requirements.
• Key Benefit: It enables knowledge workers to efficiently access and analyze data from various on-premise systems, improving productivity and decision-making.