← Cases
United StatesTechnology and telecom2025 – ongoing

Palo Alto Networks

Claude coding assistant for the engineering organization

Palo Alto Networks' developers shipped features 20-30% faster after adding a Claude coding assistant

WorkingVendor case studyVendor ↗

Problem

Palo Alto Networks' developers spent 30-35% of their time on initial development, where costly bugs often start. New developers also needed months to learn the products.

Approach

The company used Claude models on Google Cloud inside the developers' IDE for code completion, code chat and explanations. A separate AI step in the CI/CD pipeline renames variables, adds comments and generates unit tests. About 2,500 developers were onboarded, with plans to reach 3,500.

Results

20-30%increase in feature development velocity, per AnthropicVendor case study[1]
months to weeksnew developer onboarding time, per AnthropicVendor case study[1]
70% fasterintegration tasks completed by junior developers with no prior product knowledge in a pilot, per AnthropicVendor case study[1]

Takeaway

The company aimed the assistant at early development, where developers spend the most time and where bugs are costly to fix later.

Categories

Business function
IT and R&D
Innovation type
Process innovation
AI technology
Generative AI
Evidence
Vendor case study

Sources

Source summaries

[1]Anthropic1 Oct 2025

Anthropic's customer story describes how Palo Alto Networks put Claude models into its developers' IDE and CI/CD pipeline. It reports a 20-30% gain in feature development velocity, onboarding cut from months to weeks, and junior developers finishing integration tasks 70% faster in a pilot. The page carries no date; the Internet Archive first saved it on October 1, 2025.

Full details are in the original sources.

Added 4 Nov 2026 · Updated 4 Nov 2026
Open case page ↗