Jennifer Chu-Carroll

Also known as Jennifer Chu

Jennifer Chu-Carroll is a natural language processing researcher who was part of the IBM team behind Watson, the question-answering system that defeated Ken Jennings and Brad Rutter in February 2011 at Jeopardy! After an MMath at the University of Waterloo in 1991 and a PhD at the University of Delaware in 1996, she spent five years as a member of technical staff at Lucent Technologies Bell Laboratories and joined IBM in 2001. At IBM Research, she managed the Knowledge Structures group, worked on the DeepQA project from its start in 2006, coauthored the 2010 overview "Building Watson", and shared the 2013 AAAI Feigenbaum Prize with the Watson team. In 2024, she was first author of a paper on Elemental Cognition's EC AI platform, a neuro-symbolic system that pairs a logical reasoning engine with large language models. She is a research fellow at Elemental Cognition.

Now at: Elemental Cognition
Badges: Laureate
Bridge

Timeline

1991–1996

Degrees at Waterloo and Delaware

At the University of Waterloo, she earns an MMath degree in 1991, and she completes a PhD at the University of Delaware in 1996. In 1995, she and Sandra Carberry present "Response Generation in Collaborative Negotiation" at the 33rd Annual Meeting of the Association for Computational Linguistics.
By 2001

Technical staff at Bell Labs

Before joining IBM in 2001, she spends five years as a member of technical staff at Lucent Technologies Bell Laboratories. In 1999, she and Bob Carpenter publish "Vector-based Natural Language Call Routing" in Computational Linguistics.
2001–2006

Knowledge Structures group at IBM

She joins IBM in 2001. At the IBM T. J. Watson Research Center, she is a research staff member and manages the Knowledge Structures group, which applies natural language processing and AI techniques to advanced search. In 2004, she coauthors a paper on PIQUANT, an IBM Research question-answering system that the Watson team later uses as its baseline.
2006–2011

Watson and the DeepQA project

From its start in 2006, she works on IBM's DeepQA project, and in 2010 she coauthors "Building Watson: An Overview of the DeepQA Project" with David Ferrucci and others in AI Magazine. Watson, an initial application of DeepQA, defeats Ken Jennings and Brad Rutter in a two-game Jeopardy! match in February 2011.
2012–2013

A Feigenbaum Prize with Watson

She serves as general chair of NAACL HLT 2012. In 2013, she shares the AAAI Feigenbaum Prize as a member of the IBM Watson team, cited for showing that a synthesis of AI techniques can reach human-level performance in real-time factual question answering. The University of Waterloo names her a 2013 Alumni Achievement Award winner.
2020

GLUCOSE and machine comprehension

In 2020, she coauthors GLUCOSE, a large-scale dataset of implicit commonsense causal knowledge presented at EMNLP, and "To Test Machine Comprehension, Start by Defining Comprehension" with Jesse Dunietz, Dave Ferrucci and others at the 58th Annual Meeting of the Association for Computational Linguistics.
2024

The EC AI platform

In February 2024, she is first author of "Beyond LLMs: Advancing the Landscape of Complex Reasoning", which presents Elemental Cognition's EC AI platform, a neuro-symbolic system that pairs a logical reasoning engine with large language models to solve constraint satisfaction and optimization problems. In June 2024, she coauthors LLM-ARC, which combines LLMs with an automated reasoning critic and reaches 88.32% accuracy on the FOLIO logical reasoning benchmark.
As of October 2026

Research fellow at Elemental Cognition

She is a research fellow at Elemental Cognition, where she works in natural language processing.