Behavox Demonstrates its Unique Ability to Analyze Communications Data No Matter the Quality

NEW YORK – OCTOBER 27, 2020 – Behavox, the world’s only AI-based data operating platform used by firms to catch misconduct in the workplace before it causes massive regulatory fines and company crises, announced today that it successfully interpreted 50-year-old unstructured speech data from NASA’s Apollo 11 mission with its cutting-edge voice detection technology. 

Behavox conducted this analysis to demonstrate how its voice recognition and speech processing solutions are suited to seamlessly analyze any communications data, regardless of how poor, grainy or muffled their quality. In doing so, Behavox was able to validate the accuracy of its machine learning algorithms in a public setting and completely different environment to the one that has been used to test and train such tools to analyze the world of finance and enterprise business. 

The Behavox Voice Team achieved high rankings, outperforming competitors across multiple categories in the University of Texas at Dallas’ Center for Robust Speech Systems’ Fearless Steps Challenge, which provided 19,000 hours of raw analog audio from the Apollo 11 mission to participants. After extensive research, testing, and experimentation, Behavox was able to extract 1,000 hours of voice data from the recordings, successfully removing pauses, silences, background noise, and distorted speech. Behavox found that Apollo 11’s data contained the same channel distortion, incorrect grammar, and informal use of language that the company’s solution analyzes in voice data across its diverse, multinational customer base.

“The success of Behavox’s voice team in testing our technology against these historic recordings demonstrates the strength and resilience of our voice recognition and speech processing solution,” said Alexander Glasman, Chief Data Scientist, Behavox. “The parallels between the dialogue of astronauts and their operations teams on Earth and the communication between hectic traders rushing to execute a financial transaction, or a busy Zoom team meeting are remarkably similar. It’s exciting to see that the technology we’ve invested in so much works in outer space as well as it does here on Earth.”

Among the most notable findings from the analysis were that Behavox’s voice recognition technology took just 10 hours to transcribe 1,000 hours of speech, compared to what would have taken approximately 4,000 hours to transcribe manually.

Behavox’s participation in the Fearless Steps Challenge is just one of the latest developments for the fast-growing AI company. Last week the company announced the expansion of its global operations in Japan and the Nordics, and that Danske Bank has successfully implemented Behavox’s solution. Last month, Behavox announced the appointment of Microsoft (NASDAQ: MSFT) veteran Gagan Gulati as the company’s Chief Product Officer, and earlier this Summer, expanded its presence in Montréal, Québec, with a long-term lease to accommodate continued hiring of engineering, data science, and product development teams.

More from today’s Apollo 11 Announcement

Find current job openings at Behavox

Read about the latest Behavox announcements and news coverage

Read about the latest industry trends through the Behavox blog