Weights & Biases

Profile Overview

Weights & Biases (also known as W&B) is an artificial intelligence developer platform founded in 2017 by Lukas Biewald and Chris Van Pelt, the entrepreneurs who previously founded the data crowdsourcing pioneer CrowdFlower (now Appen). Headquartered in San Francisco, Weights & Biases provides the essential MLOps tools that machine learning engineers and researchers use to track experiments, manage model checkpoints, and collaborate on project development. Its core platform allows teams to visualize model training runs, compare hyperparameter sweeps, and log dataset versions, serving as the system of record for AI engineering teams. As generative AI exploded, W&B expanded its tools to include W&B Prompts, a tool for monitoring and debugging LLM inputs and outputs, and W&B Launch for orchestrating training workloads. The company's platform is a staple in AI research, utilized by leading organizations like OpenAI, Toyota, and Nvidia to train the world's largest foundation models. In October 2021, Weights & Biases secured $135 million in a Series C funding round led by Felicis Ventures and Bond Capital, bringing its total funding to $250 million at a valuation of $1 billion, establishing W&B as the premier MLOps standard.

Flagship Offerings

Vetted AI Models