Cambridge speech recognition technology startup Speechmatics is seeking up to 20 new recruits in a rapid scale-up.

It recently launched Universal Time Alignment – a language-independent forced-alignment service to match words in text files to their counterparts in audio files. The technology accurately and automatically delivers improved content discoverability, in any language, and the company is leveraging an Arabic version.

Speechmatics boasts that it provides some of the world’s fastest and most accurate speech recognition technologies, based in the cloud and leveraging the latest advances in deep learning neural networks. It already supports a number of large companies as well as SMEs.

The technology was pioneered by Dr Tony Robinson 30 years ago at the University of Cambridge and is now pushing the boundaries in speaker-independent automatic speech recognition – claiming world-beating accuracy levels.

Cambridge has become a world-leader in speech recognition technology – evidenced by Amazon’s swoop for Evi and Apple’s more recent acquisition of VocalIQ.

Dr Robinson started his first company (SoftSound) in 1995, which was acquired by Autonomy in 2000; he subsequently led the speech recognition group and was a key architect of Autonomy’s success. 

Since leaving Autonomy in 2006, he has founded or been involved in a large number of speech-based startups – most notably SpinVox, where he built a fully automated voicemail-to-text system.

Speechmatics says it can can extract audio from almost any common file format. Alongside generating a transcript, Speechmatics also provides speaker recognition (diarisation), punctuation and capitalisation.

The company has begun to employ world-leading technology developed by the Qatar Computing Research Institute to help transcribe Arabic broadcasts and audio files into text and subtitles. It can transcribe modern standard Arabic as well as four major Arabic dialects – Egyptian, Levantine, North African and Gulf Arabic.

Speechmatics recently recruited Neil MacDonald as chief revenue officer. He has enjoyed a long and successful career in the speech-tech industry having served at senior management levels within major players including Nuance, SVOX, iSpeech and HP. 

More recently he led the B2B sales team at SwiftKey, the latest UK tech-innovator to be acquired by one of the US mega-techs (Microsoft paid $250 million).

MacDonald led the sales effort on some of the most significant advances in speech-enabled product to come to market, from the likes of Google, Nokia, Samsung, Motorola and Sony – and through his work with the Android platform added the voice to over 600,000,000 Android based devices.

• PHOTOGRAPH SHOWS: The team at Speechmatics in Cambridge