NOVI, Mich.—ENCO has unveiled Raptor, a cloud-based live streaming captioning encoder that injects the speed, power and reliability of real-time AI capabilities into the streaming workflow.
Raptor represents the convergence of pure software-defined infrastructure with AI and automated captioning workflows, providing embedded captions in an IP video stream, the company said.
Introduced at IBC2025, Raptor offers a simple, straightforward way to embed CEA-608 closed captions into a live program stream. That process incorporates the AI-based workflow from ENCO’s enCaption solutions to generate captions automatically with high accuracy. Users can also use outside cloud captioning sources.
Requiring no hardware infrastructure, Raptor is consistent with the shift broadcasters are making to move their video workflows to the cloud. Quick to deploy, Raptor offers a powerful new captioning option for live events, it said.
“With no physical encoder to maintain, customers are assured a cost-efficient and agile solution that is quick to configure and deploy,” said ENCO president Ken Frommert. “Raptor represents a new threshold in flexible, cost-efficient live and automated captioning, and we’re excited to present its capabilities to our customers and partners at IBC.”
Raptor was developed as a fully software-defined architecture to allow for customization at any point along the chain to best match the workflows and standards in use. Instead of SDI signals via fixed cabling, the solution employs SRT, RTMP or HLS video streams at the front end. As Raptor’s architecture can be hosted on a variety of public or private clouds, captions can be injected at nearly any stage, offering tremendous flexibility for deployment and operation. Upon engaging and regenerating the stream, captioning text is inserted and the outbound stream is made available to the customer, the company said.
Raptor offers the accuracy of the company’s enCaption universe and ensures compatibility with its enTranslate solutions, enabling translations to dozens of languages. Users can also customize word libraries and employ corrective filters, speaker identification and the ability to recognize music, laughter, applause and crowd noise, it said.
More information is available on the company’s website.
