Research Assistant
ETH SCAI Lab · Zürich / Nottwil, Switzerland
Analyzing multimodal wearable and clinical data for decongestion monitoring in cardiac recovery.
Supervision: Diego Paez-Granados.
Research Intern
ETH LRE Lab · Zürich, Switzerland
Developed a psychologically grounded, dynamic state-aware framework for simulating diverse and evolving student behaviours beyond static personas, enabling evaluation of adaptive LLM tutors across multi-turn interactions. This work led to DySAS, accepted at REALM @ EMNLP 2026.
Supervision: Mrinmaya Sachan and Heejin Do.
Research Fellow
Microsoft Research India · Bangalore, India
Developed a privacy taxonomy and curated a dataset for context-dependent, privacy-preserving summarization. Evaluated LLMs using automatic metrics and human judgments, improving privacy adherence over GPT-4o while maintaining summary consistency, relevance, and coherence.
Supervision: Akshay Nambi.
Software Engineering Intern
Sprinklr · Gurgaon, India
Resolved memory leaks with memlab, reducing memory consumption by over 10%. Implemented server-side rendering with Next.js and GraphQL and end-to-end tests, improving load times by over 20%. Identified render blockers and improved application performance and resilience.
Research Intern
BITS Pilani · Pilani, India
Built a fundus-image annotation tool covering soft and hard exudates, microaneurysms, hemorrhages, and neovascularization. Collaborated with a hospital to curate a 3,000-image dataset. Applied image processing and custom deep-learning models to automate lesion annotation, achieving over 80% accuracy and specificity.
Supervision: Sundaresan Raman.
Software Engineering Intern
Uber · Hyderabad, India
Built an end-to-end application using React.js, GraphQL, and Go to manage credit types for over 2,000 employees. Migrated financial data to a centralized MySQL database and automated integration and end-to-end tests, reducing onboarding time by two months.
Research Intern
Mercedes-Benz · Bangalore, India
Trained customized variational autoencoders for parameter-controlled vehicle-component design generation on a proprietary dataset of over 25,000 data points. Developed comparison heuristics and a web application for researchers to generate custom designs.
Supervision: Poonam Goyal.
Summer Research Intern
Jio Platforms · Mumbai, India
Built keyword-based image collection and annotation workflows and an automated bounding-box dataset pipeline. Trained a customized YOLOv7 model on public and proprietary datasets to distinguish people from objects and barriers; successful detections achieved over 90% model confidence.
education
M.Sc. in Computer Science
ETH Zürich · Switzerland
Coursework: Probabilistic Artificial Intelligence, Large Language Models, Mixed Reality, AI in Education, Big Data, Information Security Lab, Computational Intelligence Lab, and Computational Statistics.
B.E. in Computer Science
BITS Pilani · India
CGPA: 9.83/10. Director’s Gold Medal; full tuition waiver through the Institute Merit Scholarship for consistently ranking in the top 1%.
Coursework: Data Structures and Algorithms, Object Oriented Programming, Database Systems, Theory of Computation, Operating Systems, Computer Networks, Linear Algebra, Vector Calculus, Probability and Statistics, Discrete Mathematics, Artificial Intelligence, Computer Graphics, and Human Computer Interaction.
Secondary education
South Point School · Guwahati, India
97.2%.