Researcher

Tony Robinson

Also known as Anthony John Robinson

Tony Robinson is a speech-recognition researcher and entrepreneur who built one of the first neural network-based speech recognition systems at Cambridge in 1989. After studying natural sciences and completing a Cambridge PhD, he was first author of the 1993 paper on the ESPRIT WERNICKE project, which combined hidden Markov models with neural networks for large-vocabulary recognition. In business, he formed SoftSound in 1995, founded Cantab Research, the original name of Speechmatics, in 2006, and directed the Advanced Speech Group at SpinVox from 2008 to 2010. In 2013 he coauthored the One Billion Word Benchmark paper, and by July 2017 Speechmatics, where he was chief technology officer, covered 27 languages. He is the founder of Speechmatics, which he has led for 18 years.

Now at
Speechmatics
Works
53
Citations
1,608
h-index
15

Publication numbers from OpenAlex, as of 22 October 2026.

Timeline

  1. 1981–1989

    Physics and a Cambridge PhD

    He studies natural sciences at Cambridge from 1981 to 1984, specializing in physics, then earns an MPhil in computer speech and language processing in 1985. His PhD thesis, completed in 1989, is titled “Dynamic Error Propagation Networks”, and in 1989 he builds one of the first neural network-based speech recognition systems.

  2. 1993

    The WERNICKE project

    In 1993, he is first author of the Eurospeech paper on the ESPRIT WERNICKE project, which brings together groups from Europe and the US to extend hybrid hidden Markov model and neural network approaches to large-vocabulary continuous speech recognition. His coauthors include Frank Fallside, Hervé Bourlard, Nelson Morgan and Steve Renals.

  3. 1995–2006

    Founding SoftSound

    In 1995, he forms SoftSound Ltd, a speech technology company. Autonomy invests in SoftSound in May 2000 and acquires it outright in 2006, with a view to making unstructured video and voice data easily searchable.

  4. 2006–2012

    From Cantab Research to Speechmatics

    In 2006, he founds Cantab Research Ltd, the original name of Speechmatics. The company launches its cloud-based speech recognition services in 2012.

  5. 2008–2010

    Advanced Speech Group at SpinVox

    From 2008 to 2010, he directs the Advanced Speech Group at SpinVox, a speech-to-text provider for carrier markets. Its recognition system is used more than one million times per day for a time, and Nuance acquires SpinVox.

  6. 2013

    The One Billion Word Benchmark

    In 2013, he is one of seven authors of a paper that proposes a benchmark corpus of almost one billion words for measuring progress in statistical language modeling. The best results in the paper come from a recurrent neural network language model.

  7. 2017

    Investment and 27 languages

    In January 2017, Speechmatics, where he is chief technology officer, receives investment from IQ Capital, Amadeus Capital Partners and other backers. By July 2017, a recognition system he redesigned with two colleagues around neural network acoustic and language models covers 27 languages.

  8. As of October 2026

    Founder of Speechmatics

    He is the founder of Speechmatics, which offers automatic speech recognition in 50 languages, and he has led the company for 18 years. In March 2026, Speechmatics publishes his top insights from a Reddit session in which he answered 24 hours' worth of questions from builders.

Career

  1. Since 2006SpeechmaticsFounder, Chief Technology Officer
  2. 2008–2010SpinVoxDirector
  3. 1995–2006SoftSound LimitedFounder
  4. 1991–2001University of Cambridge

Education

  1. 1989University of CambridgePhD
  2. 1985University of CambridgeMPhil in computer speech and language processing
  3. 1984University of Cambridge

Sources and credits

Photo: We have no published photo of Tony Robinson yet. Send us a free-licensed photo.

  1. Tony Robinson (speech recognition) - Wikipedia en.wikipedia.org. Checked 20 October 2026.
  2. Speech-to-text in production: what 36 years of hard lessons taught me speechmatics.com. Checked 20 October 2026.
  3. A neural network based, speaker independent, large vocabulary, continuous speech recognition system: the WERNICKE project isca-archive.org. Checked 20 October 2026.
  4. Brit neural net pioneer just revolutionised speech recognition all over again theregister.com. Checked 20 October 2026.
  5. Speechmatics - Wikipedia en.wikipedia.org. Checked 20 October 2026.
  6. One Billion Word Benchmark for Measuring Progress in Statistical Language Modeling arxiv.org. Checked 20 October 2026.
  7. Cambridge speech recognition company closes growth funding round from leading tech investors cambridgenetwork.co.uk. Checked 20 October 2026.
  8. Dr Tony Robinson | Speechmatics speechmatics.com. Checked 20 October 2026.