Our journey began with…
Previously, we were a large media company spending over $100,000 annually on transcoding live streams. However, we discovered a game-changing technology – the Nvidia Jetson Nano. After conducting experiments, we found this device to be extremely efficient for transcoding, enabling us to significantly reduce our costs to just $5,000 annually for the same volume of streams.

This breakthrough led to the development of Visioniks. We have since expanded to become a separate entity, focusing on creating eco-friendly, energy-efficient, and cost-effective solutions for video encoding, neural networks, and artificial intelligence applications.



71, Undola, Plovdiv, Bulgaria, 4000


+359 3 254 02 05

“River” is a powerful and reliable video transcoding software that has been successfully used in production for over three years, processing up to 500 video streams 24/7. Its stability and dependability make it an excellent choice for various video processing tasks, even under high loads.

Written in C++, “River” offers a REST API, making it convenient and flexible for integration into any environment. It can transcode up to 10 simultaneous full HD streams into lower bitrates, significantly saving bandwidth and reducing network load.

Moreover, “River” offers additional functionality. During video processing, you can embed a classifying neural network like YOLO, significantly expanding the software’s capabilities and using it to solve more complex tasks.

In summary, “River” is a flexible, functional, and stable software solution for real-time video transcoding that has already proven its efficiency in real-world conditions. Its capabilities make it an excellent choice for various video processing tasks.