Leeroo: Turning workflows to production-ready, e2e trainable AI systems
Leeroo empowers developers and organizations to turn complex workflows and ideas into production-ready AI systems. Our platform enables the creation of end-to-end trainable Compound Models that combine pre-trained models, tools, and data seamlessly.
Not only do we automate model building but also automate the integration of tools and data sources, accelerating the journey from concept to deployment. By abstracting complexity, we allow users to focus on solving problems, not building infrastructure, and deliver scalable, customizable, and fully private solutions.
Co-founder and Chief Scientist Background
Our Co-founder and Chief Scientist at Leeroo holds a PhD from EPFL. Previously, the team developed an LLM-based question generation and evaluation systems at Meta AI and a compact massively multilingual translation model at NAVERLABS, both with more than 0.5 million downloads, widely benefiting the AI community.
CTO Focus on Scalable AI Systems
The Co-founder and CTO of Leeroo focuses on bridging research and production to deliver scalable AI systems. With a background at BYJU'S AI Labs, the team developed BADRI, a state-of-the-art transformer-based recommendation system trained on over 30 billion interactions. Additionally, an Efficient Cached Inference Transformer stack was built that scaled the recommendation ecosystem, impacting over 10 million students. Prior to this, there was significant AI research work done at IIT Bombay.
CEO's Diverse Background
The co-founder and CEO of Leeroo had a scientific background at Meta AI, Linkedin AI, the University of Geneva, and Idiap Research Institute. With a PhD in machine learning from EPFL, the CEO brings a wealth of knowledge and experience to the Leeroo team.
