Kush R. Varshney

Career

Varshney completed a PhD in electrical engineering and computer science at MIT in 2010 with the thesis "Frugal Hypothesis Testing and Classification," advised by Alan S. Willsky, and joined IBM Research's Thomas J. Watson Research Center as a postdoctoral researcher that year. Varshney became a research staff member later in 2010 and took on management roles from 2016, rising to distinguished research staff member in 2020 and distinguished research scientist in 2023.⁷

Varshney was named an IBM Fellow in April 2024 and directs Human-Centered Trustworthy AI research at IBM. From 2015 to 2023 Varshney was the founding co-director of the IBM Science for Social Good initiative, and in 2019 spent several months as a visiting scientist at IBM Research – Africa.⁷¹²

Research

Varshney's research covers the theory and methods of trustworthy machine learning, including AI governance, alignment, safety, interpretable models, fairness and privacy. Varshney and team created the open-source toolkits AI Fairness 360, AI Explainability 360, Uncertainty Quantification 360 and AI FactSheets 360, and co-wrote the papers introducing AI Fairness 360, AI Explainability 360 and FactSheets.¹⁷³⁴⁵

More recently, Varshney's team built the Granite Guardian family of models for detecting risks in prompts and responses of language model applications, described in a 2024 paper. Varshney's applied work has used data science for human capital management, healthcare, olfaction, computational creativity and international development.⁶⁸¹

Teaching and writing

Varshney independently published the book Trustworthy Machine Learning in 2022, which is available online, and co-wrote AI Fairness: How to Measure and Reduce Unwanted Bias in Machine Learning with Trisha Mahoney and Michael Hind for O'Reilly Media in 2020.⁷¹

Work and contributions

• Granite Guardian Open weights

  Co-author of the Granite Guardian technical report (2024), which introduces risk-detection models for LLM prompts and responses released as open source; the ibm-granite/granite-guardian repository is under the Apache-2.0 license, and the Granite Guardian 3.0 8B model card lists the apache-2.0 license. Varshney's homepage lists the Granite Guardian model family among the work of Varshney's team.⁶⁸⁹¹

• AI Fairness 360 Open-source code

  Co-author of the 2018 paper introducing AI Fairness 360, an open-source Python toolkit for algorithmic fairness released under the Apache v2.0 license; the homepage lists it among the toolkits Varshney's team created.³¹⁰¹

• AI Explainability 360 and FactSheets Published research

  Co-author of the 2019 paper introducing the AI Explainability 360 toolkit and of the 2018 FactSheets paper on documenting AI services.⁴⁵

Timeline

1. 2010Completed a PhD in electrical engineering and computer science at MIT and joined IBM Research.⁷
2. 2015Became founding co-director of the IBM Science for Social Good initiative.¹
3. Aug 22, 2018Co-authored the FactSheets paper on supplier's declarations of conformity for AI services.⁵
4. Oct 3, 2018Co-authored the paper introducing the AI Fairness 360 toolkit.³
5. Sep 6, 2019Co-authored the paper introducing the AI Explainability 360 toolkit.⁴
6. 2020Co-wrote AI Fairness, a book on measuring and reducing unwanted bias, for O'Reilly Media.⁷
7. Apr 2020Became a distinguished research staff member at IBM Research.⁷
8. 2022Published the book Trustworthy Machine Learning.⁷¹
9. Apr 2024Named an IBM Fellow at IBM Research.⁷
10. Dec 10, 2024Co-authored the Granite Guardian paper on risk detection models for LLM applications.⁶

Other documented roles

Roles the sources document that are not confirmed as current.

• Founding co-director (2015–2023), IBM Science for Social Good²¹

Honors and recognition

• IEEE Fellow · Institute of Electrical and Electronics Engineers, 2024¹⁷
• IEEE Signal Processing Society Industrial Innovation Award · IEEE Signal Processing Society, 2023⁷
• Science and Innovation Management Breakthrough (for AI Fairness 360) · Falling Walls Science Summit, 2022⁷
• Best Social Good Paper Award · ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 2014⁷
• NSF Graduate Research Fellowship · National Science Foundation, 2004⁷

Where they work

• IBM 29 people in the wiki

Related people

• Aleksandra MojsilovićAlso at IBM
• Michael HindAlso at IBM
• Ruchir PuriAlso at IBM
• Kate SouleAlso at IBM
• Marina DanilevskyAlso at IBM
• Jacob SteinhardtSafety and alignment · Interpretability
• Su Lin BlodgettEvaluation and benchmarks · Safety and alignment
• Stephen CasperSafety and alignment · Evaluation and benchmarks
• Beth BarnesEvaluation and benchmarks · Safety and alignment
• Aylin CaliskanSafety and alignment · Evaluation and benchmarks