Why We Invested in Oumi
Making AI truly open
Kahini Shah |
In a very short time, AI has begun transforming science, industries, and society itself. Yet, today’s AI landscape is dominated by a handful of large tech companies and well-funded startups that closely guard their models, training data, and methodologies. This concentration of power and lack of transparency pose significant challenges for companies seeking to deploy AI safely and researchers who want to advance the field.
That’s why we’ve invested in Oumi, which is creating a genuinely open source platform and community for developing and tuning foundation models. We are excited to be investors in Oumi’s $10M Seed round.
The exceptional technical leadership of CEO Manos Koukoumidis first drew us to Oumi. As the former leader who bootstrapped and led the Google Cloud’s PaLM efforts and then Gemini’s alignment/safety, Manos brings deep expertise in coordinating large-scale AI teams and infrastructure development. And he has assembled an impressive team of engineers from Google, Apple, Snap, Meta, and other companies to pursue this mission with him.
We are also impressed with the community of advisors and angel investors that Manos and his team are building around them, which includes researchers such as Ruslan Salakhutdinov from Carnegie Mellon, Georgia Gkioxari from Caltech, and Sergey Levine from UC Berkeley among many others.
Oumi aims to create an open source platform that enables end-to-end foundation model research collaboration and development across data curation, synthesis, pre-training, post-training, evaluation, and deployment. The platform allows enterprise customers to customize foundation models for specialized tasks and data using a fully flexible and efficient enterprise-grade platform that they can trust.
The market opportunity is substantial. Enterprise AI spending is growing rapidly, with nearly half of enterprises expressing interest in moving to open source solutions to reduce costs and vendor lock-in. Oumi is well-positioned to capture this demand by offering enterprises greater flexibility, data privacy, and cost efficiency compared to closed AI platforms.
Oumi is structured as a Public Benefit Corporation and puts AI safety at the core of its mission. By fostering collaboration between academia and industry, Oumi aims to ensure this transformative technology develops in service of society.
At Obvious, we look for companies that combine technical excellence with world positive impact. Oumi is a perfect example of this, working to democratize AI development while prioritizing safety and transparency. We’re proud to partner with Manos and the team as they build an open and transparent future.
