Built for Guidewire and Gosu.
Fluent in everything else.

Developers can do this inside their usual Git-based workflows. They can reply to CARE’s review threads on a merge request or mention CARE in a new comment, just as they would with a regular user. CARE responds in context, offering explanations, alternative solutions, code examples, and updated versions of their Gosu code and unit tests.

Yes. CARE can generate new Gosu code, refactor existing code, and create or update unit tests based on the changes in your pull request. This AI-powered assistant helps developers implement fixes and best practices faster, while ensuring that the team retains full control over what is finally merged.

Yes. You can enrich CARE’s knowledge by providing your own data, whether structured or unstructured, such as internal coding guidelines, business rules or project documentation. This enables CARE to understand your specific business context and apply your custom best practices alongside standard Gosu and Guidewire guidelines.

No. CARE does not collect or store source code or customer data. Optional analytics on feedback to CARE’s threads can be enabled only to help improve the service. The connection to the LLM is encrypted, and sensitive data can be redacted using Google DLP to help you meet strict security and compliance requirements.

CARE can be integrated into CI/CD workflows used by teams working with GitHub, GitLab, Azure DevOps and Jenkins. It supports automated code review and quality checks as part of the existing development process, helping teams catch issues before they move further through the pipeline.

CARE is delivered as a plug-and-play JAR file, which can be executed locally or integrated into an existing pipeline. The initial setup, during which you define parameters such as the merge request ID, repository key, project base URL and access token, usually takes around 2–5 minutes. After that, CARE can automatically participate in reviews for every new pull request.

CARE is ideal for teams working on Guidewire- and Gosu-based systems, especially in large or business-critical insurance projects. It brings the most value during active development, when AI-assisted code reviews can reduce the burden of frequent changes on review workloads. Beyond code checks, CARE acts as an AI-powered semantic reviewer — helping designers and architects assess intent, patterns and system impact. This reduces manual effort, enforces Guidewire and Gosu best practices, and accelerates delivery across complex programmes.

No. CARE augments senior reviewers. It automates repetitive checks, enforces Guidewire and Gosu best practices, and provides AI-driven architectural and semantic insights that would normally require input from senior reviewers. Final approval remains with the team, resulting in faster AI-assisted reviews and more consistent, high-quality code across Guidewire and Gosu projects.