Skip to content

Conversation

@TheCoder2010-create
Copy link

Added comprehensive documentation for packaging RAG systems with KitOps

This PR introduces complete documentation covering the workflow for packaging and deploying Retrieval-Augmented Generation (RAG) applications using KitOps and Jozu Hub Registry.

What this documentation covers:

RAG systems combine large language models with external knowledge retrieval to deliver contextually accurate and up-to-date responses. This guide walks users through leveraging KitOps to package these complex AI systems as portable, reproducible ModelKits that can be stored in Jozu Hub registry.
Key benefits:

  • Simplifies the complete lifecycle management of RAG applications
  • Covers packaging of models, datasets, and code dependencies
  • Enables versioning and distribution of RAG systems as standardized ModelKits
  • Provides step-by-step guidance for both packaging and deployment workflows

This documentation empowers users to efficiently build, share, and deploy production-ready RAG systems with confidence.

Add comprehensive documentation for packing RAG systems with KitOps.
This guide covers the complete workflow for packaging and deploying
RAG applications using KitOps and Jozu Hub Registry.

Signed-off-by: Manav Sutar <sutarmanav557@gmail.com>

This documentation provides a comprehensive guide for packaging Retrieval-Augmented Generation (RAG) systems using KitOps and Jozu Hub. RAG combines large language models with external knowledge retrieval to provide contextually accurate and up-to-date responses. KitOps enables the packaging, versioning, and distribution of these complex AI systems as portable, reproducible ModelKits stored in Jozu Hub registry.

### 1.2 What is KitOps?
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I don't think we need to re-introduce KitOps for integrations page. If you look closely at the other integration page docs - the layout is pretty straightforward: quick overview, pre-requisites, instructions.

We don't have to reintroduce KitOps, Jozu Hub etc.

And generally since KitOps is an open source project, we need to probably start by showing this can be done with any remote registry - like Docker Hub, Jozu Hub etc.

Refer - https://kitops.org/docs/integrations/kubeflow/#prerequisites

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This given structure is probably more well suited for a blog rather than docs. For docs you can refer to other integration guides, and re-write it

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Also I guess the need for integrations page, is primarily to show a specific tool integration rather a broader conceptual guide like with RAG systems. Maybe we can ask @gorkem to see if we can have a dedicated how-to guide section where we can have more generic guides packaging things liks RAG systems, AI Agents etc. as I think this guide shouldn't be in the integrations section

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks for the feedback! You're absolutely right. I'll restructure the documentation to match the standard integration page format:
Quick overview - Brief description of what RAG is and why you'd package it with KitOps
Prerequisites - Required tools and setup
Instructions - Step-by-step guide for packaging and deploying

I'll also:
Remove the redundant KitOps/Jozu Hub introductions
Lead with registry-agnostic instructions (showing it works with Docker Hub, Jozu Hub, or any OCI-compliant registry)
Follow the same structure as the Kubeflow integration page you referenced

Will update the PR shortly with these changes.


This documentation provides a comprehensive guide for packaging Retrieval-Augmented Generation (RAG) systems using KitOps and Jozu Hub. RAG combines large language models with external knowledge retrieval to provide contextually accurate and up-to-date responses. KitOps enables the packaging, versioning, and distribution of these complex AI systems as portable, reproducible ModelKits stored in Jozu Hub registry.

### 1.2 What is KitOps?
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This given structure is probably more well suited for a blog rather than docs. For docs you can refer to other integration guides, and re-write it


This documentation provides a comprehensive guide for packaging Retrieval-Augmented Generation (RAG) systems using KitOps and Jozu Hub. RAG combines large language models with external knowledge retrieval to provide contextually accurate and up-to-date responses. KitOps enables the packaging, versioning, and distribution of these complex AI systems as portable, reproducible ModelKits stored in Jozu Hub registry.

### 1.2 What is KitOps?
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Also I guess the need for integrations page, is primarily to show a specific tool integration rather a broader conceptual guide like with RAG systems. Maybe we can ask @gorkem to see if we can have a dedicated how-to guide section where we can have more generic guides packaging things liks RAG systems, AI Agents etc. as I think this guide shouldn't be in the integrations section

@TheCoder2010-create
Copy link
Author

That's a great point. @shivaylamba You're right that this feels more like a conceptual "how-to" guide rather than a tool-specific integration.

@gorkem - Would it make sense to create a dedicated "How-to Guides" section in the docs? We could use it for broader conceptual guides like:

Packaging RAG systems
Packaging AI Agents
Other common AI/ML patterns and workflows

This would keep the integrations section focused on specific tool integrations (like Kubeflow, MLflow, etc.) while providing a home for these more generic packaging guides that demonstrate KitOps capabilities across different use cases.

Updated documentation for packaging RAG systems with KitOps, including prerequisites, project structure, Kitfile creation, and best practices.
Copy link
Author

@TheCoder2010-create TheCoder2010-create left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I have updated the document with a more practical approach rather than just theoretical concepts.

@amisevsk amisevsk requested a review from gorkem February 3, 2026 17:09
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants