Participated · Series A · Mar 2022

Synthetaic offers Rapid Automatic Image Categorization (RAIC), a computer-vision platform that automates analysis of large, unlabeled satellite imagery and video by finding where a single example appears across datasets. The product targets use cases in defense, geospatial analytics, video security, drone monitoring and AI prototyping, and the company highlights unsupervised learning to reduce reliance on human annotation. Synthetaic demonstrated RAIC by locating the 2023 high-altitude balloon in satellite images from a simple sketch, which helped attract commercial and defense interest. Customers and partners cited in the article include the U.S. Air Force, AFWERX, The Nature Conservancy, CNN and Planet Labs. The company plans to use new funding to accelerate commercialization of its computer-vision technology and to nearly double headcount to about 80 employees by the end of the year. Synthetaic was launched in 2019 and emphasizes applications where scarce or complex data are a bottleneck to AI adoption. Synthetaic is an end-to-end AI company whose flagship RAIC (Rapid Automatic Image Categorization) platform eliminates the need for time-intensive human labeling and enables rapid object labeling, AI modeling, and solution deployment. RAIC analyzes large, multidimensional imagery datasets and lets both experts and non-experts build and run detection AI models in minutes. The company combines unsupervised and generative synthetic data with an end-to-end pipeline to rapidly train and deploy models that match or outperform traditional hand-labeled approaches. Synthetaic’s solutions have been applied in healthcare (Michigan Medicine), geospatial AI (Climate TRACE), defense (US Air Force), and conservation (National Geographic), among other industries. The company will use new funding to scale RAIC, grow its team, and expand its cloud and generative AI technologies while supporting strategic partnerships. Synthetaic recently moved into a new office in downtown Delafield and is hiring engineers and developers with applied machine learning and data science experience. Synthetaic builds synthetic image datasets to train and validate AI, combining 3D artists and modelers with generative adversarial networks to produce affordable, scalable training data. The company generates photorealistic models (for example, weapons or animals) and uses adversarial networks to produce large volumes of varied images. It validates performance by testing models trained on synthetic images against real-world data. Synthetaic offers an end-to-end service so customers do not need in-house AI expertise. Early projects include partnerships with Save the Elephants (tracking animal populations) and the University of Michigan (brain-tumor identification). The company is led by founder and CEO Corey Jaskolski, who previously worked with National Geographic and a 3D media startup; Jaskolski was recently named National Geographic Explorer of the Year. Synthetaic has raised a $3.5M seed and has now raised $4.5M in total; it is part of Lupa and Betaworks’ Betalab program.