Flower Labs

Flower provides a friendly, open-source framework for building and deploying federated learning systems. It allows users to easily integrate federated capabilities with existing machine learning projects using popular frameworks like PyTorch and TensorFlow. The platform supports enterprise-grade deployment and research into decentralized foundation model training.

Traction 10/10

Hamburg, Germany · HQ

Founded 2023

Artificial Intelligence
Data & Analytics
Developer Tools
Enterprise Software
Software Only

Updated 5 months ago

Funding

Raised to date

$23.6MRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.

Funding rounds are not available yet.

Founders

4 founders

Danny H.

Employee

Aaron Sisto

Employee

Niki Pezeshki

Employee

Chong Shen Ng

Employee

Product

Problem

Traditional machine learning approaches require centralized datasets, which raises privacy concerns and limits the ability to leverage data from distributed sources. Existing federated learning solutions often lack flexibility in terms of supported frameworks and deployment environments.

Solution

Flower is a federated learning framework designed to unify machine learning, analytics, and evaluation across diverse workloads, ML frameworks, and programming languages. It enables secure collaboration on decentralized data, allowing organizations to train and evaluate models without compromising data privacy. The platform supports a wide range of frameworks, including PyTorch, TensorFlow, and Hugging Face, and can be deployed across cloud, mobile, and edge environments. Flower facilitates the transition from research to production by providing a scalable and user-friendly infrastructure for federated learning projects.

Target Audience

The primary audience includes machine learning researchers and practitioners, data scientists, and organizations seeking to leverage federated learning for privacy-preserving model training and evaluation across distributed datasets.

Features

Framework-agnostic design supporting PyTorch, TensorFlow, Hugging Face, and more
Scalable architecture capable of handling tens of millions of clients
Compatibility with cloud, mobile, and edge deployment environments (AWS, GCP, Azure, Android, iOS, Raspberry Pi, Nvidia Jetson)
Support for federated learning, analytics, and evaluation
Easy-to-use Python API requiring as little as 20 lines of code for a full federated learning system
Platform independence, ensuring interoperability across different operating systems and hardware platforms