PhD Researcher
Siddhant Dutta
Cultural Evolution | Quantum Computing | AI & Neuroscience
I research at the intersection of quantum computing and artificial intelligence, developing novel algorithms for privacy-preserving machine learning, quantum architecture search, and sustainable AI systems.
Collaborations
Research Experience
Working with leading institutions worldwide on cutting-edge research in quantum computing, AI, and machine learning.
DecIRD & UMR Espace-Dev
Multimodal LLMs | Fact-Checking Research
Dec 2024 - Sept 2025
Awarded France Excellence Charpak Summer Training Scholarship 2025 with the French National Research Institute for Sustainable Development to combat climate misinformation leveraging multimodal fact-checking through sustainable LLM techniques.
AugPurdue University
Quantum Computing | Privacy-Enhanced ML
Aug 2024 - Jul 2025
Working with SECQUOIA on Fully Homomorphic Encryption, Mixture of Experts, Multimodal Federated Learning, and Quantum Computing to improve privacy in FL with minimal performance trade-offs.
JulCapgemini
Quantum Agents | Quantum Error Propagation
Jul 2024 - Jul 2025
Conducted study on integration of Quantum AI with Agentic AI frameworks, publishing research on periodic & diminishing convergence properties of data poisoning & error propagation in QML.
JulyStanford AIM Lab
Clinical Informatics | Generative AI
July 2024
Developed a Clinical Large Language Model integrated into Stanford AIM Lab's cloud-based tutorial for teaching clinical generative AI, Python programming, and EHR data analysis.
JunNew York University Abu Dhabi
Quantum Machine Learning
Jun 2024 - Dec 2025
Researched with eBrain, Center for Cyber Security & Center for Quantum and Topological Systems on Quantum Architecture Search and Reinforcement Learning.
OctIBM Quantum & MIT-IBM Watson AI Lab
Qiskit Advocate | Graph Quantum AI
Oct 2023 - Present
Conducted research on drug discovery using Quantum Graph Attention Neural Networks and Quantum Vision Transformers for image super-resolution.
MayUniversity of Southern Denmark & TalTech
Trustworthy & Responsible AI
May 2023 - Dec 2025
Researching Neural Scaling law, Multi-task Learning & LLM4SUM for Smart and Sustainable Urban Mobility. Leading Research at SDU Quantum Hub.
Academic Output
Publications
Research publications in top-tier journals and conferences including IEEE QCE, NeurIPS workshops, and Springer Nature journals.
Total Citations
...
h-index
...
i10-index
...
Metrics automatically synced with Google Scholar profile
Career Path
Experience & Education
Education
Nanyang Technological University (NTU)
Doctor of Philosophy (Ph.D.) in Computer Science
Artificial Intelligence & Neuroscience Track
Advisor: Dr. Mengmi Zhang
University of Mumbai - SVKM's Dwarkadas J. Sanghvi College of Engineering
Bachelor of Technology in Computer Engineering
Honors in Intelligent Computing
Professional Experience
AI & Data Systems Contractor
Poznan University of Economics and Business
- •Developed containerized FastAPI service for asynchronous job market data collection using jobspy library and Playwright for browser automation
- •Built Streamlit dashboard with LangGraph and Gemini API for dynamic job description analysis
Machine Learning Intern
CRIS - Ministry of Railways, Government of India
- •Designed a hierarchical masked LSTM for multi-class complaint classification leveraging Tensorflow with dynamic subtype gating via a custom layer, enforcing class-conditional output spaces, attaining a 94.92% type and 89.62% subtype accuracy on a blind test-set of 100k real-world complaints with millisecond inference times on CPU.
- •Devised a scalable NLI-based mislabeling detection pipeline correcting manual data annotation errors, boosting effective subtype accuracy by +18.7%.
External SDE Advisor
Gaiahub - Project Liepaja
- •Collaborated with Latvian Government on real-time emissions monitoring platform for Liepaja city
- •Curated public dashboard and city officials' tools for data-driven urban planning with GDPR compliance
Summer ML Intern
Infiheal
- •Integrated LLMs into AWS Chatbot for personalized mental health support, boosting engagement by 25%
- •Utilized AWS Cloud services, Docker & Kubernetes for scalable deployment
Featured Work
Projects
Selected projects showcasing expertise in machine learning, quantum computing, and full-stack development.
Featured Projects
FeaturedFreesound Audio Tagging
Multi-label Audio Classification
Built deep learning models that classify sounds in audio clips into different categories. Processed audio data and transformed it into visual representations using mel spectrogram images. Utilized EfficientNetV2B2 CNN architecture to improve model performance and reduce computational complexity.
FeaturedMercari Price Prediction
Deep Learning Regression Case Study
Predicted product listing prices given attributes like Brand, Category, Description, Name, Shipping, and Item Condition. Implemented LGBM and Neural Networks with Bayesian search and ensemble methods. Achieved Kaggle score of 0.412, equivalent to 39th rank on the leaderboard.
FeaturedProductVision
Amazon ML Challenge · Vision-Language Models
Won 4th place All India Rank by building an end-to-end ML system to automatically extract product specifications (weight, volume, dimensions, etc.) from e-commerce images. Achieved an overall F1 score of 0.749. Benchmarked multiple Vision-Language Models across MMBench-V1.1, DocVQA(test), and OCRBench, identifying Qwen2-VL-Instruct-7B as the best performer for spatial reasoning and OCR. Applied Activation-aware Weight Quantization (AWQ), reducing model size from 7.1B to 2.64B parameters and cutting inference time from 0.2s to 0.04s per image. An image enhancement pipeline further improved F1 by 7%.
FeaturedCurateSage
Personalized Course Recommendation System
Developed a course recommendation platform leveraging heterogeneous Graph Neural Networks with PyGeometric to deliver context-aware suggestions. Implemented Edge-Aware GNN architecture with SAGEConv layers achieving 0.03s CPU inference latency.
FeaturedStockWatch
Stock Recommendation Interface
Stock recommendation & tracking application with DeepFace authentication. Ensemble of ML techniques with BERT for sentiment analysis, CNN-LSTM for forecasting, and hybrid recommendation using DGCNN and LightFM.
All Projects
Railway Mitron
Multilingual Railway Assistant
Developed platform for Ministry of Railways incorporating WhisperX (STT), ElevenLabs-TTS, and MBART translator for swift speech-to-speech translation. Integrated Gemini-based chatbot handling queries and map integration, employing IndicNER for location detection. Implemented IVRS for station authorities.
Correctify
Text Correction Engine
Developed unified platform utilizing ML models for Multilingual Input, Auto-Completion (BERT), Auto-Correction (Flan-T5), and OCR technologies. Comprehensive solution for diverse data entry and correction needs.
CarGPT
Price Suggestion Website
Developed in Datahack hackathon (Top 10). Used VGG16 to identify dents in car images. Sparse MLP with Sparse Hstacked Tfidf Embeddings + OHE Variables to suggest optimal car prices. Extracted metrics from images and scraped online data to enhance market value estimation.
Authentica
Creating Digital Identity
Developed in ACM LOC 5.0 Hackathon. Simplifies offline verification and document submission for government documents in India. Used PaddleOCR and government APIs for document authenticity. Built comprehensive backend with FastAPI and collaborative ML approach for document classification.
Julia Set & L-Systems
Fractal Visualization
Created Multi-Julia Set using JavaFX & JavaSE toolkit taking complex number and power as input to create raster of pixel intensities determining Julia set (Fractal) form. L-System implementation taking String and diverging angle as inputs. Implemented OOP practices with inheritance.
Microsoft Malware Classifier
Malware Detection System
Implemented ML models including Random Forest & XGBoost on 350GB Microsoft malware dataset containing 9 classes. Hyperparameter-tuned XGBoost + Bigram byte file + ASM + pixel intensity. Used GCP-based Jupyter notebook and achieved test log loss of 0.00793.
AI-Powered Wealth Management
Automated Financial Planning System
Winner of Datathon 2023 at DJSCE, India. Automated financial planning and investment management system. Normalized time-series data using Fast Fourier Transform. Achieved RMSLE score of 0.002 with ARIMA model for stock suggestions. Integrated government investment schemes and insurance plans.
BAGAN-GP
Drift Aware Enhanced Balancing Augmented GAN-GP
Implemented Balancing Augmenting GAN with Gradient Penalty for driver drowsiness detection using NTHU-DDD dataset. Effectively addressed data imbalance and improved classification performance with Instance Hardness Threshold Algorithm.
Recognition
Awards & Honors
Recognition for research excellence, hackathon victories, and academic achievements.
Global Winner - QuantumAI Hackathon (GenQ)
QAI Ventures
Presented at Singapore Week of Innovation & Technology (SWITCH) 2025; Golden ticket to Asia's first QuantumAI accelerator
France Excellence Charpak Summer Training Scholarship
French Government
Recognized among the top 2 scholars out of 35 selected
All India Rank 4 - Amazon ML Challenge 2024
Amazon
Out of 75,000+ applicant pool
IEEE Quantum Week 2024 / Classiq MEGA Challenge
2024IEEE
Top 5 Special Acknowledgement
International 1st Runner Up & Audience Choice Award
202513th Annual NYUAD Hackathon for Social Good
International Year of Quantum (IYQ) 2025
International Winner - Quantum Federated Summer Hackathon
2024Zaiku Group
Prize: $2,000 USD
National Winner - Smart India Hackathon (SIH)
2023Government of India
36-hour Hackathon, 200,000+ participants, Prize: ₹100,000
National Winner - Bajaj Finserv HackRx 4.0
2023Bajaj Finserv
22,000+ participants, Prize: ₹100,000
International Winner - QWorld-QIntern Research 2023
2023QWorld
Best Paper & Presentation Award
All India Rank 253 - GATE 2024
2024IIT
Data Science & Artificial Intelligence (DA) out of 100,000+
Get in Touch
Let's Collaborate
I'm always interested in discussing research collaborations, industry opportunities, or innovative projects in quantum computing and AI.
forsomethingnewsid@gmail.com