Aleksandra Mojsilović

Career

Mojsilović spent two years at Bell Labs before joining IBM Research in 2000. In what was then IBM's mathematics department, Mojsilović built models that combined public digital information with internal data to support business decisions such as sales deployment and outsourcing, work now described as data science.¹

That work led to Mojsilović being named an IBM Fellow in 2014. Mojsilović later joined IBM Research's AI team to focus on trustworthiness and fairness, and IBM Research's February 2023 profile described Mojsilović as head of Trustworthy Machine Intelligence, with newer work extending into AI for drug discovery.¹

Trustworthy AI

With Kush Varshney, Mojsilović launched IBM's Science for Social Good program, from which IBM Research says its trustworthy AI tools emerged. In 2017 Mojsilović and team urged IBM to develop bias-mitigation software, and in September 2018 IBM Research released AI Fairness 360, an open-source toolkit of fairness metrics and bias-mitigation algorithms, listing Mojsilović among its team members.¹²³

Mojsilović also co-authored the 2018 FactSheets proposal, which calls for supplier's declarations of conformity to document AI services, and the 2019 paper introducing the AI Explainability 360 toolkit and a taxonomy of explainability techniques.⁵⁴

Work and contributions

• AI Fairness 360 Open-source code

  Co-author of "AI Fairness 360: An Extensible Toolkit for Detecting, Understanding, and Mitigating Unwanted Algorithmic Bias" (2018), which introduces the toolkit released under the Apache v2.0 license. IBM Research's 2023 profile says Mojsilović and team urged IBM in 2017 to develop bias-mitigation software, and the toolkit was released a year later; IBM Research's 2018 AI Fairness 360 announcement lists Saška Mojsilović among the AIF360 team members.³⁶¹²

• AI Explainability 360 and FactSheets Published research

  Co-author of "One Explanation Does Not Fit All: A Toolkit and Taxonomy of AI Explainability Techniques" (2019), which introduces the open-source AI Explainability 360 toolkit, and of "FactSheets: Increasing Trust in AI Services through Supplier's Declarations of Conformity" (2018).⁴⁵

Timeline

1. 2000Joined IBM Research after two years at Bell Labs.¹
2. 2014Named an IBM Fellow for work that laid the foundation for IBM's data science business.¹
3. Aug 22, 2018Co-authored FactSheets: Increasing Trust in AI Services through Supplier's Declarations of Conformity.⁵
4. Sep 19, 2018IBM Research released the open-source AI Fairness 360 toolkit, with Mojsilović on the team.²
5. Oct 3, 2018Co-authored the AI Fairness 360 paper describing the toolkit.³
6. Sep 6, 2019Co-authored the paper introducing the AI Explainability 360 toolkit.⁴

Documented roles

The sources do not confirm a current role; these are the roles they document.

• Head of Trustworthy Machine Intelligence and IBM Fellow (as described by IBM Research in February 2023), IBM Research¹

Honors and recognition

• IBM Fellow · IBM, 2014¹

Companies in their history

• IBM 29 people in the wiki

Related people

• Kush R. VarshneyConnected to IBM
• Michael HindConnected to IBM
• Hendrik StrobeltConnected to IBM
• Francesca RossiConnected to IBM
• Luis LastrasConnected to IBM
• Arthur ConmyInterpretability · Safety and alignment
• Chris OlahInterpretability · Safety and alignment
• Jacob SteinhardtSafety and alignment · Interpretability
• Duen Horng "Polo" ChauInterpretability · Safety and alignment
• Samuel MarksInterpretability · Safety and alignment