Groundlight raises $10M to make image labeling faster, more accessible
The company allows clients to create algorithms to label images using simple language and code.

Groundlight, a Seattle-based startup in artificial intelligence, has announced that it has received $10 million of funding.
Madrona Venture Group, based in Seattle, led the round of funding.
Tim Porter, Madrona's managing director, stated in a press release that Groundlight's method allows companies to use off-the shelf cameras and inexpensive equipment for quickly building and reliably operating customized models. "We're excited to be working with this brilliant team, as they build their company."
Groundlight was established in 2019. The company lets clients create simple algorithms for labeling images with code and language. Groundlight's model can be used to determine whether forklifts block aisles. For example, the warehouse might ask if forklifts were blocking an aisle in a series images.
Groundlight claims that in the past, it could take several months to label a data set, train a model, and maintain this model. With Groundlight, clients can start within minutes. Currently, the service is only available to a select group of customers. Groundlight's website lists three positions that are currently open.
Madrona has made investments in major names such as Amazon and Redfin. Greycroft Partners and Founders Co-op were among the participants in this round. Flying Fish Partners as well as EssenceVC, AscendVC, AscendVC, and EssenceVC also participated. Greycroft has also invested in Bumble, the dating app, and The RealReal, the luxury goods marketplace.
Avi Geiger, CEO Leo Dirac and Groundlight were co-founded. Geiger, according to LinkedIn, was previously the co-founder and Chief Technology Officer of PicoBrew. He spent 12 years working at Microsoft before that.
Dirac spent six years as a principal engineer at Amazon before co-founding Groundlight. His LinkedIn page shows that he was also a product manger at Google earlier in his career.