Siddhant Dutta

• Awarded the France Excellence Charpak Summer Training Scholarship (previously Charpak Lab).
• Collaborating with the French National Research Institute for Sustainable Development to combat climate misinformation leveraging multimodal fact-checking through sustainable LLM techniques, under the guidance of Prof. Amira Mouakher and Prof. Laure Berti.

• Investigating Fully Homomorphic Encryption (FHE) and its integration into privacy-preserving Machine Learning paradigms, including Mixture of Experts and Multimodal Federated Learning.
• Collaborating with Prof. David Esteban Bernal Neira (Purdue and Carnegie Mellon University) to create a pioneering proof-of-concept combining Quantum Computing and Machine Learning.
• Research submitted to ICLR and AIChE addresses key challenges in privacy enhancement with minimal performance trade-offs, emphasizing scalability and cross-domain applicability.

• Exploring intersections of algebraic geometry and cryptography through elliptic curve rational points and their implications for post-quantum cryptographic systems.
• Investigating quantum machine learning error dynamics, including periodic and diminishing convergence in poisoned datasets, to improve robustness against adversarial attacks.
• Research supervised by Eldar (CTO of Capgemini Data Insights) and Prof. Bill Buchanan OBE FRSE (Edinburgh Napier University), advancing the theoretical foundation of quantum error propagation and cryptographic resilience.

• Developed a Clinical Large Language Model (CLLM) tailored for real-world Electronic Health Record (EHR) analysis and generative healthcare applications.
• Integrated the CLLM into Stanford AIM Lab’s cloud-based tutorial designed for high school and undergraduate students to learn clinical generative AI, Python programming, and EHR data analysis.
• Collaborated with Dr. Alex J. Goodell, MD, MS, to generate synthetic patient datasets, incorporating AI-driven clinical decision support and data ethics into an interactive Jupyter Notebook tutorial.

• Conducting research in Quantum Architecture Search and Reinforcement Learning under Prof. Muhammad Shafique and Post-Doc Nouhaila Innan.
• Developed Adaptive Quantum Physics-Informed Neural Networks (AQ-PINNs) for low-carbon climate modeling, presented at NeurIPS Workshop 2024, demonstrating quantum-enhanced energy efficiency optimization.
• Designed Quantum Adaptive Deep Q-Networks (QADQN) for high-frequency financial market predictions, presented at IEEE QCE 2024, advancing quantum algorithms in economic modeling.

• Advanced quantum computational approaches in drug discovery through Quantum Graph Attention Neural Networks (QGANNs) and Quantum Vision Transformers (QViT) for image super-resolution, with applications in pharmaceutical innovation and healthcare diagnostics, targeting ICCV 2025.
• Collaborating with Dr. Khadijeh Sona Najafi on quantum many-body systems and Tensor Network optimizations.
• Contributed extensively to the Qiskit open-source library and delivered a seminar on Quantum Generative Adversarial Networks (Quantum GANs) to a global audience of researchers.

• Conducting applied research in urban mobility optimization via LLM4SUM (Large Language Models for Smart and Sustainable Urban Mobility), focusing on reinforcement learning for adaptive traffic signal control.
• Developed a robust drowsiness detection framework for mobility systems using Balancing Augmentation Generative Adversarial Networks (BAGAN) with real-time drift detection.
• Co-authored a thesis on conformal prediction for recommendation systems, earning the Best Paper Award at WI-IAT Conference 2024.

• Led an interdisciplinary study on explainable AI (XAI) in healthcare, focusing on the multi-label classification of PCOSGen datasets.
• Collaborated with Dr. Palak Handa and Prof. Nidhi Goel to benchmark deep learning models for ultrasound image analysis and structured data classification.
• Contributed to the Auto-PCOS Classification Challenge, highlighting the potential of automated diagnostic pipelines in advancing precision medicine.

Interning at the Centre for Railway Information Systems (CRIS), where I am currently contributing to the development of Speech-to-Speech Translation Systems within Edge Computing paradigms and multi-modal AI software architectures. I applied advanced algorithms to address customer concerns across India’s railway network, enhancing user experience and optimizing issue resolution efficiency. My work focused on leveraging real-time processing at the edge, improving the responsiveness of the system, and integrating multi-modal inputs such as speech & text for a seamless user interface.

Collaborating with the Latvian Government to implement a real-time emissions monitoring platform for Liepaja city, focusing on vehicle-specific emissions and air quality tracking. Contributing to developing a public dashboard and city officials’ tools for data-driven urban planning. Ensuring privacy compliance with GDPR while leveraging advanced sensors and HBEFA 4.2 methodology for precise environmental insights.

Enhanced mental health support systems by integrating Large Language Models (LLMs) into an AWS-powered chatbot, increasing user engagement by 25%. Implemented scalable solutions using AWS Cloud services, Docker, and Kubernetes, ensuring efficient application management and reliability. This work underscores the role of AI in improving accessibility and personalization in digital health interventions.

Provided expert-level guidance on diverse computer science topics, resolving over 225 user queries with a 98% satisfaction rate. Topics ranged from algorithm design and data structures to advanced systems programming and machine learning concepts. This role involved translating complex technical problems into clear, actionable solutions, demonstrating both depth of knowledge and effective communication skills.

• The curriculum covers courses such as Machine Learning, Blockchain Technologies, and Digital Signal Processing, which focus on probabilistic modeling, decentralized architectures, and frequency-domain analysis, respectively.
• Artificial Intelligence and Big Data Infrastructure courses delve into neural network optimization, distributed data systems (e.g., Apache services and Hadoop), and the scalability of machine learning pipelines. Additionally, Nature-inspired Computing explores metaheuristic algorithms like genetic algorithms and particle swarm optimization for solving NP-hard problems.
• To strengthen algorithmic and database principles, I have completed courses in Advanced Database Management Systems (focusing on concurrency control and distributed query optimization), Analysis of Algorithms, and Data Structures & Algorithms (covering algorithmic complexity, amortized analysis, and advanced data organization strategies).
• I also studied Applied Physics, Applied Chemistry, and Engineering Mathematics (I-IV), which include advanced topics like quantum mechanics, quantum computing, material science, calculus, differential equations, linear algebra, and numerical methods for scientific computational problems.

• Participated in an advanced course based on DatZ7109, “Elements of Quantum Computing and Programming”, conducted in collaboration with the Faculty of Computing, University of Latvia. The course covered foundational quantum computational principles, including superposition, entanglement, and both Clifford and non-Clifford gates.
• Gained hands-on experience with quantum programming paradigms using frameworks like Qiskit and Pennylane, and explored the development of quantum circuits for custom state preparation, along with error correction techniques.
• The curriculum delved into various quantum algorithms, including quantum algorithm curation, quantum communication protocols, and the potential of quantum systems to address computationally intractable problems.