Announcing AI Error Resolution

| 3 min. (481 words)

After months of anticipation (and invaluable input from our beta testers!) we’re so excited to officially share AI Error Resolution. We can say firsthand that this tool helps developers resolve issues with renewed speed and accuracy, using AI-powered suggestions on the root cause of errors and how to fix them. Testing has shown how effectively this feature can pinpoint the source of an error and produce the most efficient method to resolve it, accelerating the entire debugging process.

(Not a Crash Reporting customer yet? You can try out Raygun and all its features, including AI Error Resolution, for free! Start a 14-trial here. No credit card required.)

While the developer is still in charge, the LLM handles the mundane parts of error diagnosis and analysis. We’re aware that AI tools are already part of the debugging process for most developers—LLMs have been integral to our workflow at Raygun for over a year now. This feature provides richer connectivity and context to the solutions you’re already generating and brings them to where you’re already working.

AI Error Resolution auto-prompts the LLM with the relevant error context — including the stack trace, environment, and affected code — to get you better responses and surfaces them within our error monitoring tool. With a secure integration that you control, you can remove the hurdles of context-switching and the frustration of getting answers that lack awareness of your system. With built-in prompts refined by our team of engineers, you’re asking better questions to get better answers.

You can also escape the double-handling, misunderstandings, and inefficiencies that come from closed chats on individual accounts. Maintain team collaboration and visibility with multiple live user messaging, conversation history, and comment capture. Provide your people and systems with the visibility to get the job done and prevent details from getting lost in translation.

Privacy and security

While we’re here, we want to take this opportunity to make a commitment to you about user data and privacy. You’ll use your organization’s OpenAI API key to securely connect with the latest GPT models, and our usual stringent standards for data transparency and protection will apply. We want to state clearly that no model training takes place with any customer data. Read more on our approach to AI data privacy here. AI Error Resolution is powered by your AI provider, while using the prompts and parameters built in by our engineers. Raygun securely posts the prompt for inference only to your selected third-party AI integration. Read more about Raygun’s data handling policies here.

We’re really looking forward to seeing how our customers will use this feature. We’d love to hear what you think! Submit any feedback or requests to anytime, or through our forums.

There’s never been a more exciting time to get started with Raygun! Try out Crash Reporting and all its features, including AI Error Resolution, free for 14 days.