WebRTC AI solution wants to Optimize your calling experience

callstats.io has announced availability of the first artificial intelligence solution for WebRTC, called Optimize. Optimize helps improve call quality by applying AI and ML algorithms. Optimize works in the background estimating the optimal quality for each user, using insights from the previous sessions. The insights are created in real-time for each individual end-user and are based on the data collected across all callstats.io’s customers. This makes it easier and faster for real-time communication providers to expand to new geographies and markets.

“Video calling has been available for two decades and while the quality of the codecs has improved, it is still suboptimal for many common situations,” said Varun Singh, Founder and CEO of Callstats. “Common issues in video calls include low quality video often resulting in a pixelated image, freezing or clipped media, video and audio being out of sync, or a video call suddenly ending. The problems are caused by a variety of reasons, most of which can be fixed automatically before the call starts or in real-time when the call is ongoing. Optimize gives real-time communications providers immediate access to call quality insights that would otherwise take months or even years to accumulate.”

Optimize looks at the context (for example, location, network and device characteristics) for each participant in a call and provides each participant with appropriate settings that would offer the best possible audio and video quality. Each call monitored by Callstats is analyzed and the AI algorithms use insight across the complete install base to provide the most suitable configurations for each call anywhere in the world. The configurations are provided automatically in the background without any explicit action from an end-user.

Optimize provides the following features:

• Automatically Fix Common Issues Affecting Quality in a Call: Optimize estimates call quality before and during a call by serving new configurations to, for example, changing network conditions.
  
  
• Automatically Highlight Insights from the Data: Developers can focus on solving the emergent problems highlighted by Optimize. Similarly, product managers will see automatically if their changes increased or lowered success rates. The platform looks for and highlights issues in application-level data that might have gone unnoticed by the company.
  
  
• Media Quality Model: Estimating media quality for large-scale conferencing is a complex problem because it depends on many dynamic factors: such as the layout of the videos, who the active speaker is, who each participant is viewing, etc. In addition to the dynamic layouts, media quality depends on the quality of individual streams.