Introducing, Autofix™ AI

Five years ago, we introduced Autofix™ - an industry-first automated remediation engine that could fix issues in your code with a single click. Today, we're excited to announce its evolution: Autofix™ AI, powered by Large Language Models.

From Deterministic to Intelligent

Our original Autofix™ pioneered automated code fixes with its deterministic approach. While revolutionary at the time, it was limited to handling specific patterns and predetermined fixes. Autofix™ AI represents a leap forward in our mission to automate code remediation.

Where our legacy system could address perhaps 30% of detected issues, Autofix™ AI can generate thoughtful, idiomatic fixes for nearly all issues that DeepSource identifies. This dramatic expansion in coverage means less time manually fixing problems and more time building features that matter.

How It Works

Autofix™ AI combines our deep expertise in static analysis with state-of-the-art LLMs. The process remains simple from your perspective:

When an issue is detected, our system analyzes not just the problematic code but its surrounding context - including imports, related functions, and your project's coding patterns. The AI then crafts a fix that resolves the issue while maintaining the style and intent of your code.

These proposed changes are presented in a diff view, giving you complete visibility and control. Review the suggestions, approve what works, and reject anything that doesn't fit your needs. All changes are proposed via pull requests, and you maintain final authority over every change in your codebase.

Seamless Transition

For teams already using Autofix™, we've designed the transition to be frictionless. Simply visit your Team Settings page in DeepSource dashboard, navigate to the Autofix™ section, and select "Autofix™ AI" mode.

What's next?

When we launched the original Autofix™, we envisioned a future where developers could focus on creative problem-solving instead of repetitive code maintenance. With Autofix™ AI, that vision takes a major step forward.

This goes beyond just resolving more issues; it fundamentally transforms the remediation process itself. Instead of the traditional cycle of detecting and manually fixing problems, Autofix™ AI creates a collaborative experience where intelligent automation handles the routine fixes, freeing you to focus on what truly matters: building great software.

In the coming weeks, we'll be introducing various enhancements to Autofix™ AI:

1. Iterative Fix Refinement: You'll be able to provide feedback at a diff level and regenerate fixes on the spot. If the initial suggestion doesn't quite match your codebase standards or needs tweaking, you can guide the AI to create an improved version without switching contexts.
2. Bring Your Own Model: For teams with specific compliance requirements, we're adding support to connect your preferred AI models. This flexibility ensures Autofix™ AI can adapt to your organization's unique needs while maintaining the seamless experience.

Experience the future of code remediation today, and see how Autofix™ AI can transform your development workflow from detection to resolution.

Note: Legacy Autofix™ will continue to be supported for a transition period but will be deprecated in the future. We recommend migrating to Autofix™ AI for the most comprehensive remediation capabilities.