German startup Flower Labs secures €18.5M to train AI on distributed data; here’s how

Hamburg-based Flower Labs secures $20M (approximately €18.56M) in a Series A round of funding to boost the widespread adoption of federated and decentralised AI.

The company’s approach challenges traditional GPU-centric methods, offering flexibility with training data and reduced dependence on GPUs. With open-source communities for distributed AI training, Flower Labs aims to align with upcoming AI regulations.

Investors supporting Flower Labs

The investment was led by Felicis, with participation from investors such as First Spark Ventures, Factorial Capital, Beta Works, Y Combinator, Pioneer Fund and Mozilla Ventures. 

The round also saw participation from angel investors, including Hugging Face CEO Clem Delangue and GitHub co-founder Scott Chacon. These investors build and support open-source community initiatives, aligning with Flower’s mission.

Niki Pezeshki, General Partner at Felicis, says, “Flower’s novel approach to federated machine learning will make model training more secure, safer, and friendly to enterprises of all sizes.”

“Top universities and multi-national companies already use Flower’s open-source technology, and the team is committed to making federated learning accessible and efficient for a wide range of users and applications.”

“We’re confident that Flower will play a pivotal role in shaping the future of AI, and we’re proud to be part of this journey.”

Capital utilisation

Flower Labs will use the funds to enhance the open-source framework, simplifying federated AI solutions. 

With opportunities in Large Language Models (LLMs) and Generative AI, Flower’s FedGPT technology enables solutions for a broader range of use cases, and increased investment will expedite its deployment.

The Flower mission

Flower Labs’ mission is to revolutionise AI model training and usage by challenging the current centralised training standard. While traditional AI relies on large-scale data collection in the cloud, Flower advocates for decentralised alternatives like federated learning. 

In this approach, data stays at its source, ensuring privacy and only computation results are transferred. Flower Labs facilitates the safe utilisation of otherwise inaccessible training data, making AI development faster, more efficient, and privacy-centric. 

Fortune 500 companies like Samsung and Nokia Bell Labs, alongside innovators like Brave and Banking Circle, are early adopters of the Flower framework.

The Flower framework boasts over 1000 open-source projects, collaborations with industry leaders like Intel, and a community of 3000 developers. Renowned universities, including MIT, Stanford, and Harvard contribute to the platform.

The Flower Summer of Reproducibility initiative, distributing up to $100K, strengthens the virtuous cycle by incentivising groups and individuals to contribute innovative solutions to the Flower platform. The annual Flower AI Summit in London gathers hundreds of researchers and developers to discuss cutting-edge decentralised AI advancements.