Quality assurance · Production

Qualtrics engineers understand code 25% faster and save up to 30 minutes daily with Sourcegraph Cody

The problem

Qualtrics' DevX team needed to improve productivity for over 1,000 engineers who lost time leaving their IDEs to find information and manually writing unit tests, while requiring a secure AI solution compatible with their self-hosted GitLab infrastructure.

Workflow diagram · grounded in source

Developer prompts Cody in IDE

trigger

“Qualtrics engineers report having to leave their IDE to find information on the web 28% less often when using Cody”

Cody retrieves context and explains code

ai_action

“if developers know how to prompt Cody, Cody can find context and explain the code to them”

Cody generates unit test template

ai_action

“Cody can generate a template for a test”

Developer refines test via prompting

human_review

“I can prompt it to make adjustments to get the test to cover exactly what I'm looking for”

Reported outcome

After adopting Cody, Qualtrics engineers leave their IDE 28% less often, understand code 25% faster, and save 10 to 30 minutes per day — roughly 10% of development time.
Unit test work that previously took multiple developer days can now be completed in an hour.

Reported metrics

IDE exit rate28% less often

Code understanding speed25% faster

Daily time savings per engineerbetween 10 and 30 minutes of time savings per day

Development time savedroughly 10% of development time

Show all 5 reported metrics

IDE exit rate28% less often

code understanding speed25% faster

daily time savings per engineerbetween 10 and 30 minutes of time savings per day

development time savedroughly 10% of development time

unit test authoring timemultiple dev days reduced to an hour

Reported stack

CodyGitLabCode SearchClaudeGPTAWS Lambda

Source

https://sourcegraph.com/case-studies/qualtrics-speeds-up-unit-tests-and-code-understanding-with-cody

Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

After adopting Cody, Qualtrics engineers leave their IDE 28% less often, understand code 25% faster, and save 10 to 30 minutes per day — roughly 10% of development time.

What tools did this team use?

Cody, GitLab, Code Search, Claude, GPT, AWS Lambda.

What results were reported?

IDE exit rate: 28% less often; Code understanding speed: 25% faster; Daily time savings per engineer: between 10 and 30 minutes of time savings per day; Development time saved: roughly 10% of development time (source-reported, not independently verified).

How is this quality assurance AI workflow structured?

Developer prompts Cody in IDE → Cody retrieves context and explains code → Cody generates unit test template → Developer refines test via prompting.

After adopting Cody, Qualtrics engineers leave their IDE 28% less often, understand code 25% faster, and save 10 to 30 minutes per day — roughly 10% of development time.Unit test work that previously took multiple developer days can now be completed in an hour.