Quickstart
Go from zero to your first AI-powered code search in 5 minutes.
This guide walks you through setting up Enhance and running your first semantic search.
1. Create your workspace
Sign up at enhance.pm and create a workspace for your team. Choose a name that represents your organization.
2. Create a project
A project groups related repositories, issues, and documentation. Create one for your main product or service.
- Click New Project from the sidebar
- Enter a project name and optional description
- Choose a project prefix (used for issue IDs, e.g.,
APP-123)
3. Connect a repository
Link a GitHub repository so Enhance can index your code.
- Go to Project Settings → Repositories
- Click Connect Repository
- Install the Enhance GitHub App when prompted
- Select the repositories to connect
For Azure DevOps repositories, see the Azure DevOps integration guide.
4. Add your OpenAI API key
Enhance uses OpenAI for generating code summaries, embeddings, and AI features.
- Go to Workspace Settings → API Keys
- Enter your OpenAI API key
- Enhance uses
text-embedding-3-smallfor embeddings andgpt-5-minifor chat by default
Don't have an API key? Get one at platform.openai.com.
5. Run your first indexing
After connecting a repository and adding your API key, Enhance automatically starts indexing.
The pipeline processes your code through these steps:
- Clone — Shallow clone of your repository
- Discover — Walk the file tree and filter by supported languages
- Parse — Extract functions, classes, interfaces, and exports
- Resolve — Build the dependency graph between code units
- Summarize — Generate AI summaries for each code unit
- Embed — Create vector embeddings for semantic search
- Finalize — Update statistics and clean up
You can monitor progress from the project dashboard.
6. Ask your first question
Once indexing completes, go to the Chat page and ask a question about your codebase in plain English:
"How does the authentication flow work?"
"Which files handle payment processing?"
"What's the architecture of the API layer?"
Enhance searches across your code, documentation, and issues to provide grounded answers with source references.
Next steps
- Connect Confluence or Notion for documentation sync
- Set up the MCP server to use Enhance from your IDE
- Explore features in depth