Poushali Sengupta, Ph.D. Research Fellow

I am currently working as a Researcher in the RESPiRE project (Responsible Explainable Machine Learning for Sleep-related Respiratory Disorders) at the Department of Informatics and dScience Centre, University of Oslo. My work focuses on developing trustworthy and explainable machine learning methods for healthcare-related applications, particularly for long-term physiological and sleep-related data analysis.

My broader research interests include trustworthy AI, explainable AI, reliability-aware machine learning, privacy-preserving AI, and uncertainty-aware intelligent systems. Through interdisciplinary collaboration across computer science, healthcare, ethics, and data science, I aim to contribute to AI systems that are transparent, reliable, human-centered, and responsibly deployable in real-world environments.

My research area is dedicated to Trustworthy AI, which encompasses a range of critical topics including Explainability, Distributed Machine Learning, Data Privacy, Security, and Fairness. I aim to develop AI systems that are not only effective but also transparent and accountable, ensuring that users can understand how decisions are made.

By exploring Distributed Machine Learning, I investigate ways to process data across multiple devices while maintaining privacy and security. Additionally, I emphasize the importance of fair algorithms that promote equity and prevent bias in AI applications. Through this work, I strive to build trust in AI technologies and contribute to their responsible deployment in society.

Explored trustworthy AI with Prof. Subhankar Mishra, focusing on differential privacy, fairness, and federated learning. Evaluated privacy-utility-fairness trade-offs and developed privacy-preserving models.

Formal training in higher education pedagogy with emphasis on student-centered learning, assessment, and the Scholarship of Teaching and Learning (SoTL). Pedagogical Statement.

As a Teaching Assistant for the Energy Informatics course, I supported students in applying complex concepts through structured guidance and formative assessments. I helped learners in mastering assignments focused on optimizing electricity cost strategies using Real-Time Pricing (RTP) and Time-of-Use (ToU) schemes, as well as developing predictive machine learning models for wind energy forecasting.

My responsibilities included addressing individual learning needs through targeted feedback, assessing student work effectively, and employing interactive methods to foster engagement and critical thinking.

My role at Chegg involved addressing student queries with precise and simplified explanations in a timely manner to enhance their understanding. I consistently achieved high ratings by delivering quality responses, ensuring student satisfaction and effective learning outcomes.

Thesis: "Towards Trustworthy AI: Balancing Explainability, Privacy, Security, and Fairness."

Supervisors: Professors Sabita Maharjan, Frank Eliassen, and Yan Zhang.

I have successfully published one paper and currently have three drafts in progress, all of which are progressing well. I am on track to complete my PhD thesis by early 2026.

Focused on stochastic processes, real analysis, multivariate analysis, and statistical computing. Thesis on privacy by shuffling, resulting in a publication. [Thesis Link]

Covered mathematical statistics, probability theory, and applied statistical methods.

Completed Generative AI for Synthetic Data: Applications in ML – one week course on GANs, LLMs, and privacy-preserving data generation.

Completed graduate-level courses: INF-8605-1 (Interpretability in Deep Learning) and DAT945 (Secure and Robust AI Model Development).

Participated in the PACE funded exchange program. Collaborated with the Decentralized Information Systems and Data Management group on distributed learning, edge computing, and federated architectures.

Covered contemporary energy topics including local markets, microgrids, storage systems, and energy justice, combining technical and socio-regulatory perspectives.

Represent PhD researchers nationally, contribute to AI strategy meetings, cross-university collaboration, and PhD community development.

Contribute to organizing annual interdisciplinary events; coordinate with researchers across faculties; curate sessions that reflect national priorities in data science.

Managed venue logistics, coordinated on-site setup and technical requirements for poster and oral sessions; supported program committee with scheduling and participant guidance.

Chaired four insightful sessions on EV Charging, Signal Processing, IoT, and Vehicular Communications.

Contributed to the program coordination, speaker logistics for AI in neuroscience, and participant engagement.

Supported venue management, and on-site coordination. Assisted with registration, speaker guidance, and real-time session management.

Co-led public science events; managed outreach, digital content, and cross-disciplinary engagement. Organized local bar events to make research accessible to the public.

Coordinated student-industry engagement, organized placement drives, and facilitated recruiter outreach and skill-building sessions.

Training in public communication, science writing, and presentation skills.

Course on ethical, philosophical, and societal aspects of scientific research.

Organized by EICT Academy from August 20–25, 2019.

Conducted by E&ICT Academy and IIT Guwahati from July 22–26, 2019.

Advanced Instructional School on Stochastic Processes (June 24 – July 12, 2019).

Organized by the Department of Statistics from June 1–7, 2018. National Workshop: “Mathematics in the Arena of Statistics”. Presidency University, Department of Statistics (December 12–16, 2016).

Presidency University, Department of Statistics (December 12–16, 2016).

Focused on advanced statistical methods and interdisciplinary applications.

Watch Here

Discussed applications of trustworthy AI in sensitive domains.

Explored trade-offs in AI model design.

Presented framework for hierarchical explainable AI.

Shared insights on privacy in distributed systems.

Contributed to discussions on ethical AI in energy.

Participated in seminars, volunteered in welfare and blood donation camps in India.