Skip to main content

Future Developments

We are actively iterating and refining CodeRabbit, and this is a sneak peek into the upcoming milestones and releases. We are focused on making the Applied AI review better than the median manual review. In addition to that, we are looking at a holistic user experience with various integrations and types of reviews.

Applied AI Improvements​

We are actively trying to make the application of Generative AI more useful , relevant , meaningful for the coder and reviewer journey. Our immediate focus is

Accuracy and Conciseness Enhancements​

  • Refining knowledge base context understanding
  • Implementing advanced summarization techniques
  • Implementing a knowledge base feature library

Learning Refinements​

  • We have made significant improvement of the learning Enhancing reinforcement learning based on the user feedback

New Feature Enhancements​

Expanded Integrations​

We are integrating various tool chains to enable coders and reviewers to have a consistent experience irrespective of the tools. The immediate tools would be:

  • Circle CI
  • Jenkins

Communication Tool Integrations​

Communication and the user experience of review via various communication tools are going to be key. We will start with integrations to Slack and Microsoft Teams and will be diving into the design engineering of these flows further:

  • Slack: Real-time notifications and interactive discussions
  • Microsoft Teams: Code review conversations within Microsoft ecosystem

Enhanced Review Capabilities​

These are additional capabilities that can also be reviewed in the same PR to accelerate the coder and reviewer journey. This includes pipeline failure analysis and resolution, as well as vulnerability assessment.

Pipeline Failure Analysis​

  • Automated analysis of CI/CD pipeline failures
  • AI-driven suggestions for resolving issues
  • Historical tracking of pipeline performance

Finishing Touches​

Finishing touches are about experience that often take developers time away from what they like doing best - coding. But adding finishing touches is crucial and should follow the ontology and taxonomy. We will start by looking into Docstring and expand to various areas to solve pain points for coders and reviewers.

1. DocString Review​

  • Automated checks for docstring presence and quality
  • AI-powered suggestions for improving documentation
  • Resolving Doc-string conflicts in a following PR

Disclaimer: any product roadmap features mentioned below are only meant to outline our general product direction. This documentation is for informational purposes only and may not be incorporated into any contract.