Curriculum vitae
Reverse-chronological summary of engineering experience, peer-reviewed research, verified credentials, and academic foundations.
Experience
Jul 2026 - Present
Data Scientist
Myntra · Bengaluru, Karnataka
- Contributing to the homepage merchandising team, engineering modeling components for the Infinite Scroll and Ads Ranking systems.
- Extended the ranking objective from CTR alone to funnel-depth (CVR) metrics using state-of-the-art deep multi-task ranking architectures.
- Authored a modular, end-to-end training and productionization codebase, including automated conversion to Triton runtime plan files for low-latency GPU serving.
- Calibrated output probabilities for Infinite Scroll modeling to achieve targeted conversion metrics without degrading platform guardrail metrics, validated through live A/B experimentation.
Jan 2026 - Jul 2026
Data Scientist Intern
Myntra · Bengaluru, Karnataka
- Contributed to the Ads Rank team across end-to-end modeling workflows and service re-architecture.
- Shipped a replacement ads-ranking model that delivered a statistically significant CTR lift in a live A/B test, after improving offline NDCG over both the production ranker and a heuristic baseline.
- Cut ad-vector generation compute by more than an order of magnitude by reducing embedding depth while retaining ranking quality.
- Cut near-real-time activity-generation batch latency several-fold by optimizing PySpark reads, shrinking the DAG, and increasing parallelization.
- Engineered robust, idempotent Databricks PySpark embedding pipelines with multi-phase checkpointing, atomic writes, and real-time inference simulation in training data.
- Built modular continuous training and evaluation pipelines with MLflow experiment tracking, visualization, and seamless CPU-to-GPU migration to meet strict production SLAs.
- Halved model training iteration time and automated experimentation workflows through custom orchestration scripts.
Jan 2025 - Jun 2026
Graduate Student Researcher
MIDAS Lab, IIIT Delhi · New Delhi
- Conducted M.Tech thesis research under the supervision of Dr. Rajiv Ratn Shah: 'Lost in Perception: Isolating Perceptual and Reasoning Failures in Multimodal Physics and Geometry Reasoning' (accepted at the 9th BlackboxNLP Workshop, co-located with EMNLP 2026, Budapest).
- Designed a 5-condition diagnostic framework holding reasoning prompts fixed while varying visual grounding across 6 multimodal LLMs and 5 benchmarks, isolating perception vs. reasoning failure modes at per-instance granularity.
- Engineered high-throughput inference backbone over OpenRouter API with thread-pool concurrency, sliding-window rate limiting, idempotent atomic checkpointing, and adaptive retry token wrappers across 140+ evaluation runs.
- Demonstrated that adding diagrams breaks up to 46.1% of problems models already solve from text alone, and question-aware captioning improves physics perception accuracy by up to 33 points.
- Engineered an async food-packaging annotation anomaly detection service using fine-tuned Qwen2.5-VL-72B on a 5× H100 cluster; built FastAPI REST APIs and Celery workers cutting latency 35% and sustaining 100+ req/s.
Jan 2024 - Aug 2025
Undergraduate Student Researcher
Face Recognition Research Project · Gandhinagar
- Improved face recognition model accuracy by 5% under classroom lighting and angle constraints by implementing pluggable CBAM and Shuffle Attention modules in PyTorch.
- Architected a decoupled cloud deployment with FastAPI, Celery workers, and PyTorch model caching to offload intensive deep learning compute from mobile and edge devices.
Jan 2024 - Jun 2024
Jr. Software Developer, Intern
Ishitva Robotic Systems · Ahmedabad
- Streamlined training pipelines by developing synthetic data generation utilities in C++ with OpenCV, improving waste-detection model accuracy by 4% (to 93% overall).
- Engineered GCN-based mask clustering models using the PyTorch metric-learning library to automate dataset organization and training sample synthesis for transparent objects.
Education
Aug 2024 - Jun 2026
M.Tech in Computer Science and Engineering
IIIT-Delhi, New Delhi
- Graduation CGPA: 8.08 / 10.0 (final, finalized graduation CGPA).
- Thesis: 'Lost in Perception: Isolating Perceptual and Reasoning Failures in Multimodal Physics and Geometry Reasoning' (Advisor: Dr. Rajiv Ratn Shah). Accepted at 9th BlackboxNLP Workshop @ EMNLP 2026.
- Advanced coursework: Machine Learning, Natural Language Processing, Deep Learning, Graduate Research in Systems.
Aug 2020 - Jun 2024
B.E. in Information Technology
LDRP Institute of Technology and Research, Gandhinagar
- Graduation CGPA: 8.14 / 10.0 (First Class with Distinction).
- Winner, State-Level Hackathon (SSIP Gujarat 2022) for Student Dropout Analysis (published in IEEE I2CT 2024).
- National Finalist (Runner-Up), KAVACH-2023 National Cyber Security Hackathon for Advanced ANPR & Face Recognition.
Apr 2019 - Apr 2020
Higher Secondary Certificate (H.S.C. GHSEB)
Parth School of Competition and Science, Vadodara
- Overall Score: 64% with Science & Mathematics specialization.
Apr 2017 - Apr 2018
Secondary School Certificate (S.S.C. GSEB)
Shree S.K. High School, Lunawada
- Overall Score: 92% (Distinction).
Selected Publications
| Year | Title & Venue | Link |
|---|---|---|
| 2026 | Lost in Perception: Isolating Perceptual and Reasoning Failures in Multimodal Physics and Geometry Reasoning Accepted at the 9th BlackboxNLP Workshop, co-located with EMNLP 2026 in Budapest, Hungary on October 29th, 2026. | Paper |
| 2024 | Unlocking Enigmatic Pathways: Empowering Student Dropout Analysis with Machine Learning and Energizing Holistic Investigation 2024 IEEE 9th International Conference for Convergence in Technology (I2CT), Pune, India. Indexed in IEEE Xplore. | Paper |
Certifications
Amazon Web Services (AWS) · July 2025
AWS Certified Machine Learning Engineer - Associate
ID: 5e46b311411640059b6fd0c09d906369
Verified Credential
Microsoft · June 2024
Microsoft Certified: Azure Data Scientist Associate
ID: E569DA1CD528CDA5
Verified Credential
Volunteering & Leadership
May 2025 - Present
Placement Coordinator (M.Tech CSE 2024-2026)
Placement Office, IIIT Delhi
- Official student representative of the M.Tech CSE cohort in the Institute Placement Committee.
- Liaising between student cohort and Institute Placement Cell to manage recruiter engagements, interview pipelines, and on-campus recruitment drives.
Dec 2024 - Apr 2025
Executive Committee Member (Curations)
TEDxIIITD
- Curated speaker lineups, reviewed talk scripts, and collaborated on editorial theme development for the annual conference.
Jun 2024
Technical Lead
LDRP Institute of Technology & Research
- Directed technical operations, team coordination, and computing resources for campus technical symposium.
Jan 2022 - Aug 2022
Creative Writer & Web Content Lead
Break The Barrier Hackathon
- Authored hackathon website copy, keyword SEO, host keynote scripts, participant guidelines, and code of conduct.