
We're getting rid of the last barrier to entry for building custom computer vision models: training data. Go from a JSON describing your ideal model's behavior, to a live endpoint hosting an appropriate model, in seconds. And if the model ever makes a mistake, just modify the JSON describing the edge case and it will fix itself automatically.

We're getting rid of the last barrier to entry for building custom computer vision models: training data. Go from a JSON describing your ideal model's behavior, to a live endpoint hosting an appropriate model, in seconds. And if the model ever makes a mistake, just modify the JSON describing the edge case and it will fix itself automatically.
What they do: Automates training-data generation and deployment for custom computer-vision models from a JSON specification
Founded / HQ: Founded 2022; New York, New York, United States
Stage / Funding: Seed stage; last reported round Feb 2024; total funding reported as $630,000
Team size: Reported 1–10 employees
Computer vision tooling; automated training-data generation and model deployment
2022
Artificial Intelligence / Computer Vision
Seed round reported in February 2024; investor names and specific amounts obfuscated in available evidence