The initial charter with k8sgpt.ai is to create a tool for giving end-users Kubernetes SRE super powers. As of 2023 there is an interesting intersection of technology and interests that now brings us to an opportunity to create a compelling Kubernetes focused solution using generative pre-trained transformers and highly mature Kubernetes API.
To do this will be a combination of human domain knowledge, codified analysis, parsing and computation of logs and configuration by remote AI services to simplify messages back to the user. There are additional possibilities of multi-language support, definition of specific problem types and even integration into existing CI/CD based workflows.
This project will be licensed under Apache-2.0 license and open source. The project team will strive to create AI providers that have options that are a combination of proprietary & non-proprietary services.
- Create a CLI that enriches telemetry from Kubernetes to assist with SRE ( Site reliability engineering ) and personas whilst debugging and triaging issues.
- Build a project that has a sustainable set of maintainers and community so that end-users have support and assurance that they are using a project with longevity.
- Showcasing AI as a transformer of complex inputs to reduce and provide simple advice on how to assist with analysis.
- Open source the project and find an appropriate long term foundation to support continued efforts.
- Creating a generic mouthpiece for a specific AI implementation