Diptikalyan Saha

I am a Senior Technical Staff Member and Master Inventor at IBM Research, Bangalore. 

My research interests span Artificial Intelligence (AI), Natural Language Processing (NLP), Knowledge Representation, Program Analysis, Security, Software Debugging, Testing and Verification, and Programming Languages.

My current focus is on AI Testing, which I categorize into two complementary areas:

• AI4Testing – applying AI techniques to improve software testing; and
• Testing4AI – testing AI models to ensure their trustworthiness.

Within AI4Testing, my research explores API Testing, Z Testing, ABAP Testing, and Chaos Testing. In Testing4AI, I work on evaluating AI models for fairness, robustness, and explainability, as well as on broader challenges including model comparison, fault localization, and verification of AI systems.

In earlier phases of my research, I worked on developing natural language dialogue systems for cognitive interaction with both structured and unstructured data. In Software Engineering, my contributions include work on automated fault localization, program repair, program equivalence, program refactoring, program tracing and profiling, program migration and translation, API extraction, and grammar inference.

I received my Ph.D. in Computer Science from the State University of New York at Stony Brook, where I was advised by Prof. C. R. Ramakrishnan and Prof. Scott A. Smolka. I earned my B.E. in Computer Science and Engineering from Jadavpur University. Prior to joining IBM Research, I worked at Motorola India Research Lab, Bangalore, and Interra IT, Noida.

Recent News (2025-)

• Presented "Generative AI in Software Engineering" at MSRIT, Bangalore.
• Attended Nasscom Agentic AI Confluence 2025 and represented IBM at the Industry Roundtable on ‘Responsible AI in the Agentic Era’
• Paper accepted at ICSE 2026 Research Track titled "An LLM Agentic Approach for Legal-Critical Software: A Case Study for Tax Prep Software". This is a joint work with Prof. Saeid Tizpaz-Niari
• Organizing the First workshop on "AI for Software Modernization" (AISM) at ASE 2025. Check out the webpage https://aism25.github.io/aism25/
• Paper accepted at ASE 2025 Demo Track titled "Evaluating Program Coverage for Code-Model Training".
• Paper accepted at FSE 2025 Industry Track titled "Automated Testing of COBOL to Java Transformation"

Past Highlights (2024-):

• I am invited to serve as the PC Co-Chair of FSE Industry Track 2023
• I am invited as a workshop chair for ISEC 2023  
• Presented 'Testing and Debugging AI Models' at FAccT'2022 CRAFT session.  
• I am invited to serve on the PC of ICSOC'2022. 
• Presented AI Fairness on webinar organized by the Department of Telecommunication, Government of India. (Mar'22) 
• I am invited to serve on the PC of ESEC FSE  Industry Track 2022 (Mar'22). 
• Presented FAT from Industry Perspective, ACM ARCS 2022 Panel on FAT (Feb'22)
• Paper accepted at MLSys'22 titled 'FROTE: Feedback Rule-Driven Oversampling for Editing Models.' - Öznur Alkan, Dennis Wei, Massimiliano Mattetti, Rahul Nair, Elizabeth M. Daly, Diptikalyan Saha.  (Jan'22)
• Paper accepted at AAAI'21 titled 'Generate Your Counterfactuals: Towards Controlled Counterfactual Generation for Text' - Nishtha Madaan, Inkit Padhi, Naveen Panwar, Diptikalyan Saha
• I am invited to serve on the PC of ISEC 2021 
• I am invited to serve on the PC of ICDE 2021 (Demo Track)
• I am invited to serve on the PC of FAcct 2021 (Research Track)
• 3 Papers accepted at CODS-COMAD 2021
• Paper accepted at VLDB 2020 titled ATHENA++: Natural Language Querying for Complex Nested SQL Queries. 
• I am invited to serve on the PC of ICSE 2021 (Research Track)
• I am invited to serve on the PC of ESEC/FSE 2021 (Research Track)
• Our work on Data Rule creation and anomaly detection is available through CP4D
• I am invited to serve on the PC of AI4SG 2020 workshop 
• I am invited to serve on the PC of ICSOC 2021 (Research Track)
• Our work on Fairness Verification got accepted at UAI 2020. Code and Paper is available thru http://www.auai.org/uai2020/accepted.php
• Our work on Explainable Payload Data Drift in AI Systems is available through AI OpenScale in IBM Cloud
• I am invited to serve on the PC of ESEC/FSE 2020 https://2020.esec-fse.org/
• I am invited to serve on the PC of ICSOC 2020 (Research Track)
• I am invited to serve on the PC of VLDB 2020 (Research Track)
• I am invited to serve as the PC Co-Chair of ISEC 2020
• I am invited to serve on the PC of CODS-COMADS 2020 
• Our work on Bias in AI Systems is available through AI OpenScale in IBM Cloud as well as thru open-source AI Fairness 360.