Code Tool Overview
When Glean is connected to your organization’s code repositories like GitHub or GitLab, Glean Assistant gains powerful code generation capabilities that boost your software engineers’ productivity. Unlike traditional generative AI or co-pilot tools, Glean provides a unique advantage by understanding and incorporating context from multiple sources - including code, design docs, technical documentation, tickets, PRs, issues, and discussion threads.

​
Overview

By integrating with your organization’s code repositories, Glean enriches its enterprise knowledge graph to provide:

Code Search

Advanced search capabilities across current and historical code content, including pull requests and commits, helping developers quickly find relevant code examples and references.

Code Generation

Intelligent code snippet generation that aligns with your organization’s standards, accelerating feature development and bug fixes while maintaining consistency.

​
Key Benefits

Rapid Error Resolution

Quickly identify and resolve issues using a comprehensive database of code snippets, documentation, and previous solutions.

Automated Code Generation

Boost productivity with automatic generation of code snippets for both routine tasks and new feature development.

Knowledge Sharing

Facilitate effective communication of coding practices and standards across your organization.

Streamlined Onboarding

Accelerate new engineer integration with structured learning resources and codebase insights.

​
Example Queries

​
Code Context

  • “How is [system] implemented at our company?”
  • “How do I debug/solve [error]?”
  • “Help me write code to do X in [language]”
  • “Where is the code file that does X?”
  • “Help me write unit tests for [file]”

​
Process & Customer Context

  • “What is the status of this week’s backend release?”
  • “Who can I ask about [system or feature]?”
  • “How do I set up my dev environment?”
  • “Which customers are using [feature]?”
  • “What customers have raised issues this week?”

​
Usage Scenarios

​
Adding to Existing Code

1

Situation

An engineer needs to create a new evalset, working with code spread across multiple files and languages. They need to leverage existing code rather than starting from scratch.
2

Example Queries

  • “How do I add a new search evalset?”
  • “How to create a new autocomplete evalset for history suggestions?”
3

Outcome

Adding to Existing Code Example
Glean Assistant identifies relevant files requiring changes and provides explanations with adaptable code snippets.

​
Debugging Errors

1

Situation

An on-call engineer encounters a panic occurring in specific product situations.
2

Example Queries

  • “Help me debug [full panic error]”
  • “Can you review the code and help me fix the panic?”
  • “Give me a list of reviewers for this code change”
  • “Can you create a PR description?”
3

Outcome

Debugging Process
Error Resolution
Glean Assistant analyzes the stack trace, suggests fixes, and assists with PR creation.

​
Feature Status Check

1

Situation

An engineer needs to determine a feature’s implementation status and customer requests.
2

Example Queries

  • “Is [feature] enabled?”
  • “What customers have asked for this feature?”
3

Outcome

Feature Status Example
Glean Assistant compiles information from design docs, code changes, and Jira tickets, including feature flags and customer requests.

​
Code Change Impact

1

Situation

An engineer debugging an error needs to understand recent changes affecting their component.
2

Example Queries

“What recent changes could have caused errors in runtime configs?”
3

Outcome

Code Change Impact Analysis
Glean Assistant analyzes recent changes across code, tickets, and discussions to provide a comprehensive summary.

​
Implementation Guidance

1

Situation

An engineer starting a new implementation needs guidance on multi-file changes across the stack.
2

Example Queries

“Where can I add a new endpoint for QE?”
3

Outcome

Implementation Guidance Example
Glean Assistant provides file locations and required changes based on similar past implementations.

​
FAQ