Kubeflow Goals

Author:Francisco Javier Arceo | **Date:**Nov 20, 2024

We are at a critical moment in time for AI/ML. Open-source models have taken center stage, and resilient, scalable infrastructure has become critical to the success of AI. However, infrastructure alone isn't enough—tools must inspire excitement and engagement among users.

As a Kubeflow Steering Committee member, my core focus would be to drive engagement and increase adoption by improving the experience and impact of Kubeflow for end users, ensuring it's not only a robust platform but also one that people love to use. More precisely, I would focus on 6 key goals.

1. Expanding Kubeflow Adoption: increasing awareness, adoption, and engagement through collaborations with enterprises and startups.
2. Making RAG a first class priority: the AI community has settled on Retrieval Augmented Generation (RAG) as a critical component for production AI. I will work to make this a new focus area for Kubeflow.
3. Improving the User/Developer Experience for Model Developers/Users: simplifying workflows and enhancing usability to reduce barriers to entry.
4. Engaging and Empowering AI Engineers: supporting software engineers integrating AI/ML into their products. By providing intuitive workflows, thoughtful defaults, and subsets of Kubeflow's tooling, we can ensure AI Engineers are set up for long-term success as they scale their products.
5. Building a more Feature-Complete AI/ML Platform: driving development of critical features to support end-to-end AI/ML workflows.
6. Improving Documentation: ensuring comprehensive, user-friendly docs and resources for the community.
7. Strengthening the Community: fostering a vibrant, inclusive contributor and user ecosystem to support Kubeflow's long-term growth.

I believe I am qualified to help lead Kubeflow's values and structure because of my past open-source contributions and professional experience.

As a leader, maintainer, and contributor to Feast, I have helped evolve a critical component of the MLOps ecosystem, driving innovation in feature storage and retrieval. My 12+ years of experience at the intersection of data engineering, AI/ML engineering, and MLOps include technical and managerial leadership roles at Red Hat, Affirm, Fast, Goldman Sachs, and others. Across these positions, I've designed, built, and scaled AI models and platforms in production, gaining a deep understanding of user pain points and opportunities to innovate.

I bring this experience to Kubeflow with a vision of making it the leading open source platform for scalable, production-grade AI/ML. By expanding adoption, improving the user experience, and delivering a feature-complete solution, my goal is to empower MLOps teams to build a powerful AI/ML platform that enables their success and expands the global impact of AI.