October 5, 2021
Rendered.ai, a platform provider for physics-based synthetic data, has announced raising $6 million in a seed round, led by Space Capital. The company said the funding will help expand staffing and operations to scale up its platform-as-a-service (PaaS) that supports data scientists, engineers and business innovators who want to use synthetic data to investigate, validate, and expand AI-based technology and business opportunities.
AI solutions require large datasets for training, which is often a limiting factor when building AI capabilities, Rendered.ai said. Real-world datasets based on physical sensor-based collection can be expensive, have bias and gaps, or even be impossible to obtain. Simulated or synthetic datasets can power AI applications for industries such as satellite reconnaissance, robotics, medical imaging, and security applications where real-world data is limited.
“Synthetic data is an answer to the most pressing problems in AI,” said Nathan Kundtz, founder and CEO of Rendered.ai. “Many users focus only on generating static datasets. By offering access to a PaaS for iterative collaboration and data generation, we can help customers and partners adopt synthetic data as an innovative, cost-effective business capability.”
Rendered.ai provides a developer framework for synthetic data. The platform integrates scenario generation, 3D model libraries, asset management, compute management, annotation, metadata management, and analytics into a single offering to generate content to train machine learning systems, for data labeling and annotation, and to validate and improve AI accuracy.
“The utility of synthetic data is already obvious to customers, however most existing solutions are highly bespoke and unscalable,” said Chad Anderson, managing partner at Space Capital. “By providing an open platform, Rendered’s approach will enable the broad adoption of synthetic data into existing enterprise workflows.”
The Rendered.ai platform is a cloud-native experience, including a no-code configuration tool, imagery generation using high-performance compute, and microservices and APIs to characterize, integrate, and access data with enterprise systems. The platform also enables teams of data scientists and data engineers to work together to configure sophisticated pipelines for data generation to fit their business needs.
“We have seen the capability and potential fo Rendered’s platform demonstrated with customers and we believe that it will become the gold-standard for synthetic data generation,” said Anderson. “We are thrilled to join the efforts of Nathan and the team at Rendered and look forward to continuing to support their growth into geospatial intelligence markets, which are hungry for their services.”
The company’s users include leading commercial and governmental organizations, academic partnerships, and CV analytics firms. In addition to Space Capital, investors in the seed round include Tectonic Ventures, Congruent Ventures, Union Labs, and Uncorrelated Ventures.
For more details about the platform, visit the Rendered.ai website here.
Photo: Rendered.ai generated this aerial image from a dataset used to detect cranes. The Rendered.ai platform was used to engineer this dataset for effective detection using physics-based simulation and GAN-based post processing.