W

Weights & Biases

AI · San Francisco, United States · Founded 2017

Weights & Biases provides a developer-first MLOps platform designed to streamline the machine learning lifecycle. The company was founded in 2017 by Lukas Biewald, Chris Van Pelt, and Shawn Lewis. Biewald and Van Pelt had previously co-founded Figure Eight (formerly CrowdFlower), a company focused on providing training data for machine learning. Their experience led them to identify a critical need for better software and best practices in ML development, which prompted the creation of Weights & Biases. The platform offers a comprehensive suite of tools for ML practitioners, including experiment tracking, hyperparameter optimization, dataset versioning, and model performance visualization. Key features include "Experiments" for tracking and comparing ML experiments, "Sweeps" for automating hyperparameter searches, and "Artifacts" for versioning datasets and models, ensuring reproducibility. The system is designed to serve both individual researchers and large enterprise teams by facilitating collaboration, tracking model lineage, and managing the entire workflow from development to production. W&B integrates with popular ML frameworks like TensorFlow, PyTorch, and Keras. The company's business model is subscription-based, with tiered pricing that includes a free option for personal use and paid plans for teams and enterprises. Weights & Biases has attracted significant investment, raising a total of $250 million over several funding rounds. A notable Series C round in October 2021 raised $135 million, and a subsequent round in August 2023 added $50 million, bringing the company's valuation to approximately $1.25 billion. Its investors include Insight Partners, Coatue, Felicis Ventures, and Bond. The platform is used by over 700,000 ML practitioners and major organizations like OpenAI, NVIDIA, Meta, and Toyota. In March 2025, it was announced that CoreWeave acquired Weights & Biases. Keywords: MLOps, machine learning, AI developer platform, experiment tracking, hyperparameter tuning, model versioning, data visualization, LLMOps, model management, collaborative AI development, dataset versioning, AI agents, deep learning tools, model registry, AI lifecycle management, production AI, machine learning workflows, model optimization, reproducibility, MLOps platform