What challenges did Palo Alto Networks face before AI implementation?

Creating the DOR was manual and time-intensive, requiring highly skilled employees to gather information from Salesforce and internal knowledge bases. The process could take days, delaying opportunities and pulling experts away from higher-value customer strategy work.

What was the impact of the AI solution?

The automation dramatically reduced the time required to produce a DOR. Standardizing the workflow improved consistency and completeness across DORs.

Source

GCPProductionEvidence: Low40/100

Palo Alto Networks automates customer intelligence document creation with agentic design on Vertex AI Agent Engine

Use case typeDocument automationUpdated Jun 13, 2026

Palo Alto Networks, a global cybersecurity company, built an internal AI-driven workflow to automate creation of its Document of Record (DOR) for sales and pre-sales engagements. The DOR provides a standardized 360-degree customer view for sales and support teams and was previously assembled manually from Salesforce data and multiple internal knowledge sources.

Organization: Palo Alto Networks
Industry: Tech & Comms
Location: United States
Published: January 2026

Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks

Customer: Palo Alto Networks
Provider: GCP
Maturity: Production
Linked source: Google Cloud Blog

The agent is deployed on Vertex AI Agent Engine and uses Vertex AI RAG Engine with Vertex AI Discovery Engine Search to retrieve grounded context from documents and logs stored in Google Cloud Storage

Customer identity supportedSource describes one deploymentMaturity supported

Primary read

Use case focus

Showing 3 of 4

1Agentic workflow orchestration
2Document automation
3Retrieval-augmented generation

Palo Alto Networks developed an AI agent using Google's Agent Development Kit (ADK) to answer more than 140 standardized DOR questions.
The agent is deployed on Vertex AI Agent Engine and uses Vertex AI RAG Engine with Vertex AI Discovery Engine Search to retrieve grounded context from documents and logs stored in Google Cloud Storage.
A FastAPI orchestrator running on Google Kubernetes Engine sends batched requests to the agent endpoints, validates responses, manages state, and writes completed output back to Salesforce through Cloud Pub/Sub.
Gemini 2.5 Flash is used for question preprocessing, and Gemini is used to synthesize grounded answers from retrieved snippets.

Technologies

Agent Development Kit Vertex AI Agent Engine Vertex AI RAG Engine Vertex AI Discovery Engine Search Gemini 2.5 Flash Gemini Google Kubernetes Engine Cloud Pub/Sub Google Cloud Storage FastAPI Salesforce

The automation dramatically reduced the time required to produce a DOR.
Standardizing the workflow improved consistency and completeness across DORs.

Architecture

A FastAPI webserver on Google Kubernetes Engine orchestrates requests from Salesforce to two agent endpoints on Vertex AI Agent Engine. The agents use Vertex AI RAG Engine with Vertex AI Discovery Engine Search for retrieval from Google Cloud Storage documents and logs, then use Gemini for answer synthesis. The webserver consolidates results, publishes the completed DOR to Cloud Pub/Sub, and a downstream service writes it back to Salesforce.

Sources & evidence1

Evidence: Low40/100Evidence strength