weights & biases raises $45m for its machine learning tools
Weights & Biases, a company developing resources for those working in machine learning, has announced the completion of a $45 million Series B funding round.
The organization was established by Lukas Biewald, Chris Van Pelt and Shawn Lewis – Biewald and Van Pelt were previously the founders of CrowdFlower/Figure Eight (which was later acquired by Appen). Weights & Biases reports having over 70,000 users currently, working within more than 200 businesses.
Biewald (who has been known to the author since their time in college) explained that while machine learning specialists are frequently likened to software developers, “they often resemble scientists more closely than engineers.” This field involves a substantial number of experiments, and Weights & Biases’ primary offering enables practitioners to monitor these experiments. The company also provides tools for managing data set versions, assessing models, and overseeing pipelines.
“If a model is operating a self-driving vehicle and an accident occurs, it’s crucial to understand the sequence of events,” Biewald stated. “If that model was created some time ago and numerous experiments have been conducted since, it can be challenging to systematically determine what transpired” without utilizing experiment tracking.
He characterized the startup as “a leading innovator” in this area, and noted that its dedication to “the ML practitioner” distinguishes it from emerging competitor tools that focus on enterprise-level sales. He also mentioned that with the increasing adoption of machine learning, Weights & Biases sometimes encounters a “positive challenge.”
“We aren’t interested in working with organizations that are implementing machine learning simply for the sake of it,” Biewald said. “Some companies have a directive from leadership to integrate machine learning throughout the business, which I find discouraging, as it may lack meaningful impact. However, I would say that the majority of the companies we engage with are genuinely pursuing valuable applications.”As an illustration, he cited John Deere, a major agricultural company, which is leveraging the startup’s platform to consistently refine its robotic application of fertilizer – as opposed to pesticides – for weed and pest control. Additionally, pharmaceutical companies are utilizing the platform to model the behavior of various molecules.
Weights & Biases had previously secured $20 million in funding. This latest round was spearheaded by Insight Partners, with contributions from Coatue, Trinity Ventures and Bloomberg Beta. George Mathew of Insight Partners will be joining the company’s board of directors.
“I haven’t encountered a category leader in MLOps with such a strong NPS and a commitment to customer satisfaction as Weights and Biases,” Mathew commented in a press release. “I’m pleased to make my initial investment at Insight to support an ML practitioner user base that has grown sixtyfold in the past two years.”
The startup plans to allocate the new funding towards continued expansion of its teams in engineering, growth, sales, and customer support.
Anthony Ha