Hey YC! I’m Alex, founder of Dragoneye.

I want to share with y’all the launch of our new interactive Playground, which allows you to build your own custom video detection models in under 5 minutes.

The Problem:

Video detection models are incredibly useful. Analyzing security camera footage, ensuring PPE compliance, monitoring production lines.

But it’s always been hard to build video detection models. Thousands of hours of videos that need to be annotated. You train the model only to find that it doesn’t perform as well as you need it. Really cool ideas that never get to prototype, never mind released in production. Crucial features that take tens of thousands of $$ to trial and months to implement.

Our Solution:

Our new Playground allows you to build and deploy production-ready video detection models in less than 5 minutes, enabled by our powerful zero-shot models.

To get started:

1. Upload a video
2. Describe the categories of items you want to detect in your video
3. The model streams the predictions and you can preview it on your video
4. Iterate on your categories till you’re set.

Once you’re happy with your model, you can save it out and it’s live! To make it easier to integrate with your apps, we have libraries for Python and Node.js, with more languages to come.

Check out the Playground at https://playground.dragoneye.ai/ to get started. You can follow our step by step guide below as well!

https://youtu.be/2rhKBO4N7B0

We can’t wait to see what you build! 🙂