Inspiration

We understood the need for automated error detection and resolution in production systems. The idea came from wanting to create a system that could not only detect errors but also automatically fix them using AI, reducing manual intervention and downtime.

What it does

Dispatch is an automated error detection and resolution system that monitors and fixes errors. It provides real-time error monitoring with SMS communication.

How we built it

We built it using Next.js for the API routes, Datadog for log monitoring, Twilio for SMS notifications, and AllHands OpenHands API for AI-powered PR creation. The system includes automated polling to wait for PR completion and sends status updates via SMS.

Challenges we ran into

The main challenge was integrating multiple APIs (Datadog, Twilio, AllHands) and handling asynchronous operations like waiting for AI agents to complete their work. We also had to implement proper error handling and TypeScript typing throughout the system.

Accomplishments that we're proud of

We're proud of creating a fully automated error-to-fix pipeline that requires zero human intervention. The system successfully combines log monitoring, SMS notifications, and AI-powered code generation into a seamless workflow.

What we learned

We learned how to integrate multiple external APIs effectively and handle long-running asynchronous operations. The project taught us about error monitoring best practices and the importance of proper TypeScript typing in complex systems.

What's next for Dispatch

Next steps include adding support for multiple repositories, implementing more sophisticated error categorization, and adding webhook integrations for other monitoring tools beyond Datadog.

Built With

Share this project:

Updates