Larry Heck

Also known as Larry Paul Heck

Larry Heck is a speech and language researcher whose work runs from early deep neural networks for speaker recognition to the virtual assistants Cortana, Google Assistant and Bixby. After a Georgia Tech PhD in 1991, he joined SRI, where his team built large-scale deep neural network technology for speaker recognition, and in 2005 he moved into web search and advertising science. At Microsoft from 2009, he started the project that became Cortana; he founded Google's Deep Dialogue team in 2014 and led Bixby product teams in North America as CEO of Samsung's Viv Labs. He returned to Georgia Tech in 2021 and is a professor in Electrical and Computer Engineering and Interactive Computing, where his AI Virtual Assistant (AVA) Lab builds next-generation assistants. Named an IEEE Fellow in 2016, he was elected to the National Academy of Inventors in December 2024.

Now at: Georgia Institute of Technology
Born: 1963
Badges: Founder
Bridge
Boomerang
Wayfarer

Years link to their sources.

Timeline

1986–1991

Speech research and a Georgia Tech PhD

Born in Havre, Montana, he earns a BSEE at Texas Tech University in 1986, an MSEE at Georgia Tech in 1989 and a PhD there in 1991 under James H. McClellan. He builds his first speech recognizer in 1987, and his dissertation applies speech algorithms to machinery sounds and vibrations.
1992–2005

Speaker recognition at SRI and Nuance

He joins the Stanford Research Institute (SRI) in 1992, moving from acoustics research to the Speech Technology and Research (STAR) Laboratory, where he founds the SRI Speaker Recognition Group. Funded by the NSA and DARPA, his team builds large-scale deep neural network technology for speech processing for the 1998 NIST Speaker Recognition evaluation. From 1998 to 2005, he is vice president of R&D at Nuance Communications, where he deploys this deep neural network with the Home Shopping Network in 1999.
2005–2009

Search sciences at Yahoo!

In 2005, he moves to Yahoo! as vice president for search and advertising sciences, with a team that develops Yahoo! web search and advertising products. In 2008, he and Ron Brachman combine search and advertising quality with Yahoo! Research to form Yahoo! Labs. He stays at Yahoo! until 2009.
2009–2014

Founding Cortana at Microsoft

In early 2009, Satya Nadella asks him to write the long-range plan for Microsoft Search, later Bing; he then becomes chief scientist of the Microsoft Speech team and starts Project Louise, which becomes Cortana. He is named a Microsoft Distinguished Engineer in 2012 and joins Microsoft Research. His 2013 paper "Learning deep structured semantic models for web search using clickthrough data" leads to a major innovation in Microsoft's Bing search engine.
2014–2017

Deep Dialogue at Google

He joins Google Research in 2014 as a principal scientist and founds the Deep Dialogue team, which builds conversational AI technology for the Google Assistant. In 2016, he is named an IEEE Fellow for leadership in applying machine learning to spoken and text language processing.
2017–2021

Bixby and Viv Labs at Samsung

He joins Samsung in 2017, co-leads Samsung Global AI Research and creates AI research centers in Silicon Valley, Toronto and Montreal. As president and CEO of Viv Labs, a Samsung subsidiary, he leads the North America product teams for the Bixby assistant. In 2019, he leads a large-scale deployment of Transformer-based language models in the Bixby Categories launch at the Samsung Developer Conference.
2021–2024

Return to Georgia Tech

He joins Georgia Tech's School of Electrical and Computer Engineering in August 2021 as a professor, Rhesa S. Farmer Chair and Georgia Research Alliance Eminent Scholar. In 2023, his post extends to a joint appointment in the School of Interactive Computing, and he becomes co-director of the new AI Hub and the Machine Learning Center. In December 2024, he is elected a fellow of the National Academy of Inventors, and CIKM honours his 2013 web search paper with its Test of Time Award.
As of October 2026

Professor and AVA Lab director

He is a professor with a joint appointment in Electrical and Computer Engineering and Interactive Computing at Georgia Tech, where his AI Virtual Assistant (AVA) Lab works on next-generation virtual assistants. In 2026, he coauthors papers on how large language models represent emotion, including "Emotions Where Art Thou" at ICLR. In August 2026, he gives a SIGDIAL keynote on probing and controlling interactional structure in multimodal foundation models.