PeopleofSuperintelligence

Researcher

Patrick Nguyen

Patrick Nguyen is a speech-recognition researcher whose papers run from eigenvoice speaker adaptation to deep neural network acoustic models and end-to-end speech systems. After starting a company for real-time foreign exchange trading in 1998 and earning a doctorate at EPFL in 2002, he worked at Panasonic's Speech Technology Laboratory in Santa Barbara and then at Microsoft Research from 2004 to 2010, where he and Geoffrey Zweig wrote on segmental conditional random fields. By 2011 he was a research scientist at Google, where he coauthored the 2012 paper Deep Neural Networks for Acoustic Modeling in Speech Recognition and took part in sequence-to-sequence speech recognition and multispeaker speech synthesis work in 2017 and 2018. In 2019 he coauthored the Lingvo framework and the StarNet point-cloud detector. He is a research scientist at Google.

Now at
Google
Badges

Timeline

  1. 1998–2002

    A trading startup and a doctorate

    In 1998, he starts a company that provides a market-maker platform for real-time foreign exchange trading. He receives his doctorate from EPFL in 2002.

  2. 1999–2004

    Eigenvoices at Panasonic

    He works at the Panasonic Speech Technology Laboratory in Santa Barbara from 2000 to 2004. His 1999 Eurospeech paper with Christian Wellekens and Jean-Claude Junqua applies eigenvoices, a rapid speaker adaptation technique, to speech recognition in noisy environments, and "Rapid speaker adaptation in eigenvoice space" appears in IEEE Transactions on Speech and Audio Processing in 2000.

  3. 2004–2010

    Microsoft Research and segmental CRFs

    He works at Microsoft Research in Redmond from 2004 to 2010. In 2008, he contributes to the winning English-to-Chinese machine translation effort in a NIST competition. With Geoffrey Zweig, he publishes "A Segmental CRF Approach to Large Vocabulary Continuous Speech Recognition" in 2009, which proposes a segmental conditional random field framework built on acoustic detectors.

  4. By 2011

    Google Research in Mountain View

    By 2011, he is a research scientist in Google Research in Mountain View, a coauthor of about 15 granted patents, and a member of the organizing committee of ASRU 2011. In July 2011, he speaks to the SF Bay ACM chapter on segmental conditional random fields in speech recognition, work done with Geoff Zweig at Microsoft Research.

  5. 2012

    Deep networks for acoustic modeling

    In 2012, he is one of eleven authors, including Geoffrey Hinton, Li Deng and Dong Yu, of "Deep Neural Networks for Acoustic Modeling in Speech Recognition" in Signal Processing Magazine. The paper presents the shared views of four research groups on deep neural networks that outperform Gaussian mixture models on a variety of speech recognition benchmarks.

  6. 2017–2018

    Sequence-to-sequence speech at Google

    In 2017, he is a Google Brain contributor to "State-of-the-art Speech Recognition With Sequence-to-Sequence Models," a joint effort with the Speech team whose end-to-end model reaches a 5.6% word error rate on a voice search task, against 6.7% for the best conventional system. In 2018, he coauthors "Transfer Learning from Speaker Verification to Multispeaker Text-To-Speech Synthesis," a system that generates speech in the voice of many speakers, including speakers unseen during training.

  7. 2019

    Lingvo and StarNet

    In 2019, he coauthors Lingvo, a TensorFlow framework for sequence-to-sequence models that dozens of researchers already use in more than 20 papers. He also coauthors StarNet, a point-based detector for LiDAR point clouds that is competitive with or better than convolutional baselines on the Waymo Open Dataset and KITTI.

  8. As of October 2026

    Research scientist at Google

    He is a research scientist at Google, with interests in speech recognition, machine learning, natural language processing and speaker recognition. His 2012 paper on deep neural networks for acoustic modeling has more than 15,000 citations.

Career

  1. NowGoogleResearch Scientist
  2. 2017–2018Google Brain
  3. From 2011Google ResearchResearch Scientist
  4. 2004–2010Microsoft Research
  5. 2000–2004Panasonic

Education

  1. 2002EPFLDoctorate

From the biography.

  1. Google alumni have founded or led 133 other organizations, more than any other AI labIncludes Patrick Nguyen. We estimate that alumni of 17 major AI labs went on to found or lead 350 other organizations, including 226 they founded, and we trace 472 career moves directly between 23 labs since 2010.
  2. Google had the most people shaping AI every year from 2014 to 2025Includes Google. Our estimates of which organizations employed the most people shaping AI each year from 1955 to 2025, tracing the shift from university labs to companies.
  3. Google has the highest combined h-index among AI organizationsIncludes Google. We estimate that Google holds the highest combined h-index at 8,016, while universities account for 73% of the deduplicated total across AI organizations.
  4. 20% of people at companies who shape AI also hold a university role, led by Google and Stanford UniversityIncludes Google. We estimate that 539 of the 6,766 people with a known current organization (8%) hold both a university role and a company role, including 20% of people at companies and 13% of people at universities.

Colleagues

People named in this biography, and people with dated positions at the same organization in the same years.

Sources and credits

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

  1. Free ACM talk on Speech Recognition - SF Bay ACM sfbayacm.org. Checked 22 October 2026.
  2. ISCA Archive - Maximum likelihood eigenspace and MLLR for speech recognition in noisy environments isca-archive.org. Checked 22 October 2026.
  3. Patrick Nguyen - Google Scholar scholar.google.com. Checked 22 October 2026.
  4. A Segmental CRF Approach to Large Vocabulary Continuous Speech Recognition - Microsoft Research microsoft.com. Checked 22 October 2026.
  5. Deep Neural Networks for Acoustic Modeling in Speech Recognition - Google Research research.google. Checked 22 October 2026.
  6. Improving End-to-End Models For Speech Recognition - Google Research Blog research.google. Checked 22 October 2026.
  7. State-of-the-art Speech Recognition With Sequence-to-Sequence Models - arXiv arxiv.org. Checked 22 October 2026.
  8. Transfer Learning from Speaker Verification to Multispeaker Text-To-Speech Synthesis - arXiv arxiv.org. Checked 22 October 2026.
  9. Lingvo: a Modular and Scalable Framework for Sequence-to-Sequence Modeling - arXiv arxiv.org. Checked 22 October 2026.
  10. StarNet: Targeted Computation for Object Detection in Point Clouds - arXiv arxiv.org. Checked 22 October 2026.