KServe Easy Deploy: Helm-based Onboarding Experience for ML Developers

[akshaymittal143]

CNCF Project Name: KServe – Standardized Serverless ML Inference Platform on Kubernetes

Title: KServe Easy Deploy: Helm-based Onboarding Experience for ML Developers

• Description:

KServe is a powerful platform for deploying machine learning models on Kubernetes. However, its setup can be complex for newcomers, especially those without deep Kubernetes expertise. This project aims to simplify the onboarding process by developing a Helm chart and an optional CLI tool that streamline the deployment of KServe and associated components.​

Key Objectives:

- Develop a Helm chart that automates the installation of KServe along with optional components like Knative, Istio, Cert Manager, and Prometheus.​

- Create an optional CLI tool (kserve-lite) to further simplify deployment and management tasks.​

- Provide sample configurations and documentation for deploying common ML models (e.g., scikit-learn, HuggingFace Transformers).​

- Ensure the solution is customizable to cater to various user needs and environments.

• Expected Outcome:

- A publicly available Helm chart hosted on ArtifactHub or the CNCF repository.​

- An optional CLI tool (kserve-lite) for simplified user interactions.​

- Comprehensive documentation, including step-by-step guides and sample configurations.​

- Improved onboarding experience for new users, leading to increased adoption of KServe.

• Recommended Skills:

- Familiarity with Kubernetes and Helm.​

- Experience with YAML, Bash scripting, and optionally Python.​

- Understanding of machine learning model deployment processes.​

- Basic knowledge of DevOps practices and tools.

• Mentor(s): Akshay Mittal ([email protected]) https://www.linkedin.com/in/akshaymittal143/

• Upstream Issue (URL): A proposal issue will be created in the KServe GitHub repository to discuss and track the development of the Helm chart and CLI tool: https://github.com/kserve/kserve/issues

[rvs1729]

Hi @akshaymittal143, This is a great idea! I’d love to contribute and be involved in bringing it to life.

[KKDevOps07]

Hi @akshaymittal143, i am interested in contributing this, can you get back if the issue is still open.

[KKDevOps07]

Hi @akshaymittal143, where do you want me to start, can you guide me.

[akshaymittal143]

Hi @akshaymittal143, where do you want me to start, can you guide me.

Please slack me on CNCF, lets discuss next steps!

[diya-18]

Hi @akshaymittal143 I want to contribute. I'd be obliged if you could guide me through it.

[akshaymittal143]

Created a issue on Kserve: https://github.com/kserve/kserve/issues/4560

[diya-18]

thanks akshay!

On Mon, Jun 30, 2025 at 9:12 PM Akshay Mittal @.***> wrote:

akshaymittal143 left a comment (cncf/mentoring#1412) https://github.com/cncf/mentoring/issues/1412#issuecomment-3019691194

Created a issue on Kserve: kserve/kserve#4560 https://github.com/kserve/kserve/issues/4560

— Reply to this email directly, view it on GitHub https://github.com/cncf/mentoring/issues/1412#issuecomment-3019691194, or unsubscribe https://github.com/notifications/unsubscribe-auth/BDHQPWI74CGUEO7DG4MOD433GFLE7AVCNFSM6AAAAAB26ZPCUCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTAMJZGY4TCMJZGQ . You are receiving this because you commented.Message ID: @.***>

[nate-double-u]

Hi, I'm sorry I missed this issue earlier. Thanks for your interest, but the CNCF LFX program here is only for CNCF projects.