AI product teams

# Put a fleet on   your next release.

Investigate recurring failures, develop changes and test them against the product you already build.

Built around your project

## More approaches.  Working at once.

Bring the implementation, failed traces and the fixes you have tried. Agents can explore tool behavior, context and application logic at once, sharing reproductions and candidate changes.

01 

## Agent improvement

Investigate traces and develop code changes for recurring failures.
[Put a fleet on it ](/use-cases/agent-evaluation/) 

02 

## Performance optimization

Try different implementations against the workloads that cost you time.
[Put a fleet on it ](/use-cases/performance-optimization/) 

03 

## Technical exploration

Build working tests of the approaches you are considering.
[Put a fleet on it ](/use-cases/technical-diligence/) 

## Work that returns  to your team.

Your existing agent brings back code, tests and findings from the fleet. Review the changes in the project where you need them, then choose what to ship or investigate next.
[See work from real fleet runs ](/examples/) 

## Put a fleet on your project.

256 agents. Four hours. $1,000 for your first run, with model inference, computers and fleet operation included.

[Explore your first fleet ](/first-fleet/) [Ask your agent what to run ](/evaluate/)