The National University of Singapore campus at dusk
Portrait of Siva Sai
Google Scholar

Siva Sai

Postdoctoral Research Fellow National University of Singapore

I work on machine learning systems pulled apart across many parties, and on what it takes to keep them accurate, private and trustworthy once they are. Currently: distributed quantum machine learning.

Publications Research
60
Journal and conference papers
2,848
Citations
25
h-index
40
i10-index

Counts from Google Scholar, including 37 papers in IEEE journals and magazines and 12 in Transactions.

Research

One question runs through all of it: when a learning system is split across hospitals, vehicles, devices or quantum processors that cannot fully trust each other, what breaks, and how do you fix it?

Quantum machine learning for distributed settings

My current work at NUS studies the adversarial robustness of quantum models trained across separate parties, and quantum generative modelling. At Melbourne I built APS-pQFL, a personalised quantum federated learning method that uses soft masking and circuit-level sensitivity analysis to keep accuracy up on highly skewed non-IID medical data, the regime where a single shared model degrades worst.

Privacy-preserving distributed learning

Split learning, federated learning and incremental learning for clinical data that is not allowed to leave the institution holding it, with the client selection, feature selection and lightweight-client problems that come with Internet of Medical Things deployments.

Blockchain for accountable systems

Ledger-backed protocols for parties that have to cooperate without trusting one another: e-prescription management, medical record sharing, anomaly detection in connected-vehicle platoons, and traceability in physical supply chains.

Generative AI, capability and risk

Taxonomies and evaluations of diffusion and language models across healthcare, security and consumer systems, together with machine unlearning as a way to take training data back out of a deployed model.

Wireless sensing and applied ML

Wi-Fi channel state information for human activity recognition, fall detection and hand hygiene monitoring, plus diffusion-guided augmentation for vision-based accident detection in vehicular networks.

Structuring emerging fields

Where an area is moving too fast to have a shape yet, I build one: taxonomies and architectural frameworks for quantum AI in mission-critical systems, for generative AI in the quantum era, and for machine unlearning. Each one sets out the open problems I then go after.

About

I am a postdoctoral research fellow at the National University of Singapore, where I work on the adversarial robustness of distributed quantum machine learning and on quantum generative modelling. I completed my PhD at BITS Pilani in December 2025 under Prof. Vinay Chamola, on the confluence of blockchain and AI technologies for next-generation healthcare solutions, and spent the closing months of 2025 at the University of Melbourne as a visiting researcher with Prof. Rajkumar Buyya, working on distributed quantum machine learning. Before the PhD I read Computer Science and Economics at BITS Pilani, finishing with a CGPA of 9.03 out of 10.

The through-line across my work is distribution and trust. I have built split learning and federated learning frameworks for Internet of Medical Things deployments, blockchain-backed protocols for prescription management and connected-vehicle anomaly detection, and most recently APS-pQFL, a personalised quantum federated learning method for heterogeneous healthcare networks. The recurring technical problem is heterogeneity: clients that hold very different data, run on very different hardware, and cannot be trusted to behave. APS-pQFL answers it on the quantum side with soft masking and circuit-level sensitivity analysis, personalising each client's circuit instead of forcing a single shared model onto all of them. On the classical side I have built a client selection method that chooses which hospitals to train with rather than averaging over everyone, a feature selection algorithm for incremental learning on IoMT streams, and CSITime, a privacy-preserving Wi-Fi sensing model for human activity recognition that needs no cameras or wearables. The work appears in IEEE Transactions on Consumer Electronics, IEEE Transactions on Intelligent Transportation Systems, IEEE Internet of Things Journal, ACM Transactions on Internet Technology, Future Generation Computer Systems and Neural Networks, and has been cited roughly 2,800 times.

Positions

Feb 2026 to present

Postdoctoral Research Fellow

National University of Singapore

Adversarial robustness of distributed quantum machine learning, and quantum generative modelling.

Sep to Dec 2025

Visiting Researcher

University of Melbourne

Distributed quantum machine learning with Prof. Rajkumar Buyya. Developed APS-pQFL for privacy-preserving medical diagnostics across IoMT-enabled institutions, co-authored a Morgan Kaufmann book chapter on quantum federated learning, and contributed to work on quantum-integrated high-performance computing.

May 2023 to Jul 2024

Research Internship, Generative AI for Immersive Systems

Edinburgh Napier University, with Prof. Amir Hussain

Led an international team working on generative models for immersive and metaverse systems: which architectures hold up under the latency and compute budgets of real deployments, and how they fail. Two journal papers.

Dec 2022 to Jan 2023

Research Internship, Blockchain and AI for Healthcare

National University of Singapore, with Prof. Biplab Sikdar

Led an international team on secure and scalable healthcare solutions, and on consumer electronics technologies for the metaverse. Three papers in IEEE journals and magazines.

Jan to Jul 2021

Research Internship, Distributed Ledgers for Connected Vehicles

University of Texas at San Antonio, with Prof. Heena Rathore

Implemented TangleCV, a distributed ledger framework for connected vehicles, from scratch in Python and evaluated it on the SwRI dataset under parasite chain attacks and sensor faults.

May to Jun 2019

Summer Data Science Intern

Couture.AI, Bangalore

Feature extraction and matching over a large apparel image database using SIFT, SURF, BRIEF, histogram and LBP methods in OpenCV, with a Flask API for a dress-object recognition model.

Selected publications

Nineteen papers that represent the range of the work, first-authored technical contributions first. Filter by area, or see the complete list on Google Scholar.

2026

APS-pQFL: adaptive parameter sparsity for personalized quantum federated learning in heterogeneous healthcare networks

S. Sai, S. Kumar, V. Chamola, M. Gurusamy, R. Buyya

IEEE Internet of Things Journal

Impact factor 8.9

Under revision

2026

Quantum machine learning for cybersecurity: a taxonomy and future directions

S. Sai, I. Goyal, S. Sharma, S. H. Manuri, V. Chamola, R. Buyya

Software: Practice and Experience

Impact factor 3.5

2026

Quantum federated learning: architectural elements and future directions

S. Sai, A. Sawaika, P. Singh, R. Buyya

Book chapter in Federated Learning: Foundations and Applications, Morgan Kaufmann, pp. 325 to 343

Four further papers appear in workshop proceedings at EMNLP, EACL and AAAI. See the full list on Google Scholar

People I work with

Prof. Rajkumar Buyya

University of Melbourne

Distributed and quantum machine learning, quantum-integrated high-performance computing.

Prof. Biplab Sikdar

National University of Singapore

Secure and scalable healthcare systems, consumer electronics for the metaverse.

Prof. Vinay Chamola

BITS Pilani, doctoral advisor

Blockchain and AI for next-generation healthcare.

Prof. Amir Hussain

Edinburgh Napier University

Generative AI, cognitive computation, machine unlearning.

Prof. Heena Rathore

Texas State University

Distributed ledgers and anomaly detection for connected vehicles.

Teaching

  • Deep Learning, teaching assistant BITS Pilani, Aug to Dec 2021. Prepared and assessed assignments and projects for a class of 80.
  • Database Systems, teaching assistant BITS Pilani, Jan to May 2021. Ran lab components and doubt-clearing sessions for a class of 200 sophomores.
  • Mentoring I have led teams of junior students and international collaborators through survey and systems papers in all five of my research areas, from first outline to camera-ready.

Recognition

  • Outstanding Research Article Award EEE Department, BITS Pilani, for the blockchain-enabled split learning framework published in IEEE Transactions on Consumer Electronics.
  • Merit Cum Need Scholarship BITS Pilani, every semester of the undergraduate programme. Awarded to the top 3 percent of students.
  • HASOC-Dravidian-CodeMix 2020 First place in Task 1 and Task 2a, second in Task 2b, on multilingual offensive speech detection in Manglish, Tanglish and code-mixed Malayalam.

Contact

Email is the quickest way to reach me. I am always glad to hear about collaborations in distributed machine learning, trustworthy AI, quantum computing or health informatics.

  • Email
  • ScholarGoogle Scholar profile
  • Based inDepartment of Electrical and Computer Engineering, National University of Singapore