It support · Production

Sourcegraph case studies listing: AI-assisted code search, Cody, and Batch Changes across engineering teams

The problem

Engineering teams struggle to understand, navigate, fix, and automate code changes across large multi-repository codebases spanning multiple code hosts, often relying on teammates and spending months on tasks that could take days.

Workflow diagram · grounded in source
1
Multi-repo code navigation
trigger
“find and navigate code across multiple code hosts without relying on teammates”
2
AI code understanding
ai_action
“Qualtrics uses Cody to speed up code understanding and write unit tests in minutes”
3
Agentic AI development
ai_action
“AI agents with the right context can act as a force multiplier for developers”
4
Automated large-scale changes
output
“reduced the time for large-scale code changes by 80% using Sourcegraph Batch Changes”
Reported outcome

Engineering teams using Sourcegraph report measurable improvements in developer productivity, time savings, employee satisfaction, and the ability to address security vulnerabilities and large-scale code changes faster.

Reported metrics
Time for large-scale code changes80%
Developer productivity impactmeasurable impact of developer productivity, time savings, and employee satisfaction
Repository analysis timedays instead of months
Reported stack
SourcegraphCodyBatch ChangesUniversal Code SearchPerforceGitHub
Source
https://sourcegraph.com/case-studies/we-are-thorn
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

Engineering teams using Sourcegraph report measurable improvements in developer productivity, time savings, employee satisfaction, and the ability to address security vulnerabilities and large-scale code changes faster.

What tools did this team use?

Sourcegraph, Cody, Batch Changes, Universal Code Search, Perforce, GitHub.

What results were reported?

Time for large-scale code changes: 80%; Developer productivity impact: measurable impact of developer productivity, time savings, and employee satisfaction; Repository analysis time: days instead of months (source-reported, not independently verified).

How is this it support AI workflow structured?

Multi-repo code navigation → AI code understanding → Agentic AI development → Automated large-scale changes.