Sagnik Dakshit

Sagnik Dakshit, Ph.D, ACUE
Assistant Professor

Kennesaw State University

(Previously) Assistant Professor; The University of Texas at Tyler

Dr. Dakshit's research focuses on building trustworthy, explainable, and human-centered AI systems for high-stakes domains such as healthcare and education. I develop methods that move beyond accuracy-driven optimization to address interpretability, reliability, and real-world deployment.

He obtained his Ph.D. in Computer Sciences from The University of Texas at Dallas , USA, specializing on Intelligent Systems and B.Tech from WBUT, India. During his Ph.D., he has worked on various research projects with Robert Bosch, Seagate Technologies, HP Machine Learning Lab, Nokia Bell Labs. Prior to his Ph.D he has worked on software development projects with IBM, and Tata Technologies.

Dr. Dakshit serves as a reviewer for the Journal of Healthcare Informatics Research, IEEE Transactions on Multimedia, Multimedia Systems (Springer), ACM International Conference on Multimedia Retrieval, NIPS, ICLR, ICHI and ACM Information Hiding and Multimedia Security. He was part of the Program Committee for the International Conference on Healthcare Informatics 2020 Special Session on COVID-19.

Integrated Intelligence Lab

The Integrated Intelligence (II) Lab develops human-centered intelligent systems that are explainable, reliable, and deployable in real-world environments. The lab integrates machine learning, applied informatics, and interactive human–AI systems to ensure accountability and trust in high-stakes decision making.

Trustworthy and Explainable AI

Designing AI models whose reasoning, uncertainty, and failure modes can be meaningfully interpreted by domain experts. The lab treats explainability as a core design principle, not a post-hoc visualization, enabling safer deployment in clinical and educational systems.

Applied Informatics and Data Quality

Investigating how data quality, representation, and provenance shape intelligent behavior. This work includes explanation-guided data curation, bias auditing, and reliability analysis across multimodal datasets to support ethical and generalizable AI pipelines.

Human-in-the-Loop Multimodal AI

Developing interactive and agentic AI systems that integrate language, perception, biosignals, and context while remaining responsive to human feedback. These systems emphasize collaboration, trust calibration, and adaptive reasoning in immersive and learning environments.

Latest Updates

January 2026: Journal paper acceptance in MDPI Education Sciences “Toward Explainable Smart Learning Systems: The PEARL Framework for Ethical, Fair, and Transparent AI in Education”.
December 2025: I am happy to announce that I will be starting at KSU Spring 2026 as an Assistant Professor of Information Technology.
September 2025: We won the Best Research Contribution Award at ECML-PKDD RHCML 2025!
September 2025: Two poster presentations and a podium presentation at TACCSTER 2025.
September 2025: New paper alert in IEEE Multimedia.
July 2025: Our work on "Virtual Reality-based Doctor-Patient Interaction Simulation Using Large Language Models" " has been selected for podium presentation RUSH University AI Symposium in Chicago Oct 3rd.
July 2025: Grateful to announce we have received $175K support from NSF to advance ECG diagnosis translation.
July 2025: New research paper acceptance at RHCML ECML-PKDD, 2025.
June 2025: New Book Chapter in Artificial Intelligence, Academic Integrity, and the Internationalization of Higher Education: Navigating Innovation and Ethics. Springer Publishers, 2025
March 2025: New research paper acceptance at IEEE ICHI 2025.

Selected Publications

“CEFEs: A CNN Explainable Framework for ECG Signals”, B. M. Maweu1, S. Dakshit1, R. Shamsuddin, and B. Prabhakaran, Artificial Intelligence in Medicine, Volume 115 (102509), May 2021. https://doi.org/10.1016/j.artmed.2021.102059 (Joint First Author)
"Generating Healthcare Time Series Data for Improving Diagnostic Accuracy of Deep Neural Networks," B. M. Maweu, R. Shamsuddin, S. Dakshit and B. Prabhakaran, IEEE Transactions on Instrumentation and Measurement, https://doi.org/10.1109/TIM.2021.3077049 2.
“Bias Analysis in Healthcare Time-Series (BAHT) Decision Support Systems from Meta-Data”, S. Dakshit, S. Dakshit, N. Khargonkar and B. Prabhakaran; Journal of Healthcare Informatics, 2023
S.Dakshit, “A Discussion on Potential of RAG in Computer Science Higher Education”; ACM SIGITE, 2024
Sagnik Dakshit, Balakrishnan Prabhakaran, "Ambient Intelligence in Immersive Realities", IEEE Multimedia, 2025
N. Balasubramania, S.Dakshit, “A Comparative Study on the Responsible Use of Public LLMs for Self-Diagnosis ”; RHCML ECML-PKDD, 2025

Media Coverage

UT Tyler to launch A.I. teaching assistant this fallUT Tyler researchers developed AI tool to enhance learning by answering student questions using professor-uploaded content, ensuring accurate information.

UT Tyler professor describes how AI is reshaping job market“With the proliferation of AI and large language models like ChatGPT, we have seen a lot more interest from companies in terms of how do you build a chatbot?” says Dr. Dakshit.

UT Tyler team receives $174,000 grant to help A.I. understand EKG readingsDr. Sagnik Dakshit said AI tools often struggle with rare heart conditions, and it can be hard for doctors to trust the results.

ETTIC, UT Tyler encourage AI partnerships with local businessesSmall business leaders learned artificial intelligence is within reach at Friday’s meeting of the East Texas Technology and Innovation Coalition. Sagnik Dakshit, computer science professor at the University of Texas

WATCH: UT Tyler asst. professor studying use of ‘A.I.’ for medical self-diagnosis“The journey towards reliable AI-assisted self-diagnosis is ongoing, and it is one that must be navigated with both hope and caution.”

UT Tyler students show off innovation, creativity at Capstone CompetitionThe atrium of UT Tyler’s Soules College of Business transformed into an arena of innovation during the Student Capstone Race Competition — Unlocking the CRK Experience on Nov. 22. Sponsored

Presentations and Talks

""Functional Movement (FMOVE) Tele-Screening Application"; Southern Biomedical Engineering Conference; Shreveport, LA September 2024
"Abstaining ECG Classifiers Through Explainable Prototypical Spaces"; IEEE International Conference on Healthcare Informatics; FL June 2024
"Investigation Of Augmentation Methods For Deep Learning ECG Classification"; International Conference on Artificial Intelligence in Medicine, Utah July 2024
“CVAE-based Generator for Variable Length Synthetic ECG”, International Conference on Healthcare Informatics; TX, USA, 2023
“Twelve Lead Double Stacked Generalization for ECG Classification”, International Conference on Healthcare Informatics; TX, USA, 2023
“Core-set Selection Using Metrics-based Explanations (CSUME) for multiclass ECG”, International Conference on Healthcare Informatics; MN, USA, 2022
Keynote on “Improving Healthcare Time-Series Deep Learning Models” at Current Research in Engineering and Technology (ICCRET-2021), Kolkata, India, Nov 2021.
The Present and Future of ML & AI at Calcutta Institute of Engineering and Management, 2021.
STEM Mentoring Career Panel – ” How I ended up in Technology, why I am passionate about it, and how it impacts our lives.” Thomas Elementary School, Plano, TX, 2019
A peek into the world of Artificial Intelligence organized by STEM Professional Series (My Passion for Science). Microsoft Theatre, 2020, Stonebriar Mall, Frisco.
Invited Guest Mentor for Final Year Engineering Projects at Calcutta Institute of Engineering and Management, 2021