Pre-doctoral researcher · Indian Institute of Science
I work on getting reliable behavior out of systems that are noisy, large, and hard to reason about end‑to‑end — across quantum computer architecture and large-scale anomaly detection.
I am a pre-doctoral researcher at the Indian Institute of Science (IISc) working on quantum computing systems — how quantum programs are compiled, planned, and executed reliably on real, noisy hardware — toward the goals of India's National Quantum Mission. Alongside this runs an applied second thread: anomaly detection for fraud and AML with NPCI, whose UPI network clears over 700M transactions a day.
I came to research through an Integrated M.Sc. in Mathematics at NIT Rourkela and two years building production data and machine-learning systems in industry. The combination of mathematical foundations and real systems at scale is what draws me to questions where quantum computing and applied machine learning meet real workloads.
Quantum computers promise to expand what is computable — but today's machines are noisy, and coaxing a trustworthy answer out of them is a full-stack problem, from how a circuit is compiled to how its results are certified. My work spans that stack, toward quantum computation reliable enough to use — a goal at the core of India's National Quantum Mission.
Full-stack co-design of the quantum stack: noise-adaptive compiler design and per-phase fidelity attribution (my HBR decomposition), together with cross-platform benchmarking of real NISQ hardware — treating compiler, runtime, and device as one system.
From quantum search and optimization algorithms down to execution planning and shot-budget feasibility — sound resource estimates, and certifying when a precision target is actually reachable, so a result carries a guarantee rather than a hope.
An applied thread on the same reliability question — graph neural networks for fraud and AML detection at production scale, and hybrid quantum–classical models on hard real-world data.
Accepted IEEE Quantum Week (QCE) 2026
Under review Q-SET 2026 — 6th International Workshop on Quantum Software Engineering & Technology, IEEE Quantum Week (QCE) 2026
Under review 33rd IEEE International Conference on High Performance Computing, Data & Analytics (HiPC 2026)
Preprint A hybrid quantum-classical neural network reaching 0.943 AUC on highly imbalanced credit-card fraud data.
Technical report An exposition and implementation of Grover's unstructured search.
A simulation-free, closed-form criterion that reads compiled gate counts and device calibration to choose how to run a quantum search query: whole, cut, or factorized.
View on GitHub →Studying akrasia-like behaviour in large language models — when models "know better" but act against it.
View on GitHub →Experiments with quantum neural networks and the role of random-projection / parameterized circuits as model layers.
View on GitHub →Detection of diabetic retinopathy using quantum computing — Runner-up, Flipkart GRID 5.0.
View on GitHub →DREAM Lab, Dept. of Computational & Data Sciences · Prof. Yogesh Simmhan
Systems research spanning quantum computing (architecture & noise-adaptive compilation) and GNN-based anomaly detection at scale. Led the HBR Decomposition work (under review, QCE 2026).
Next Generation Data Management · Gurugram, India
Databases and architecture at scale; built a GenAI-powered data-architecture documentation system (LLM pipelines) cutting manual overhead ~60%. Early promotion.
New Age – Special Projects · Mumbai, India
Fine-tuned BERT / open-source LLMs on enterprise data for real-time transaction classification (~99% accuracy).
Hybrid Quantum–Classical ML · Advisor: Prof. Snehashish Chakraverty
Designed a hybrid quantum–classical neural network (2-qubit parameterized circuits as Keras layers) for credit-card fraud detection. Observing that transpilation fidelity moved model accuracy is what motivated my later compiler-fidelity work.
Odisha, India
Relevant coursework: quantum computation, quantum mechanics, probability & statistics, stochastic processes, numerical analysis, graph theory, and number theory.
I'm applying to PhD programs for Fall 2027. I'm most interested in making quantum computation trustworthy, and in reliable learning on large, noisy data and I'd be glad to talk about either.
A few moments from conferences, labs, and competitions along the way.
