New York University, NY, NY. May 2024
Master of Science, Computer Engineering, GPA: 4.0/4.0
Coursework: Machine Learning, Probability & Stochastic Processes, Deep Learning, Big Data, Applied ML in Finance, Data Structures, Java, Internet Architecture & Protocol.
Indian Institute of Technology Kharagpur, India. May 2018
Bachelor, and Master of Technology, Aerospace Engineering.
Coursework: Neural Networks, Programming & Data Structures, Innovation Management.
Skills
Languages: C++, Python, SQL, Java, Visual C#, FORTRAN, R.
Soft Skills: Problem Solving, Leadership, Agile Development, Collaboration, Adaptability.
Work Experience
Graduate Research Assistant, DICE Lab, New York University (NYU). Nov 2023 – present
- Enhanced lung cancer outcome prediction using decision trees (XGB, Catboost) over neural networks (TabPFN) on right-censored tabular data achieving 7% higher accuracy for survival rates with 25% faster training time. [Github]
Data Science Intern, LOCOMeX Inc. May 2023 – Aug 2023
- Engineered supply chain & local-content analyzer tool using XG Boost & CatBoost models for clients, achieving 99.2% accuracy in predicting waivers/exceptions under Buy America Act using 25-years of historical data. [Github]
- Reduced cost and manual efforts by 90% by building Retrieval Augmented Generation (RAG) pipeline using LangChains and LLMs for extracting ESG data from contracts, reducing 80 man-hours/week.
- Led a team of 4 interns, to conduct vector search similarity analysis on text embedding with Pinecone, improving contract opportunity categorization accuracy by 40% and retrieval speed by 75%. [Github]
Graduate Research Assistant, Syracuse University (SU). Feb 2021 – Jul 2022
- Primary developer of Python and C++ modules in ROS which implements the Finite-Time Stability concept to forecast unknown disturbances to an Unmanned Aerial Vehicle (UAV) with an accuracy of 95.3%.
- Focused on debugging & linear algebra methods to predict unknown disturbances using numerical methods. [Paper][Github]
Software Developer, Airbus India. Feb 2019 – Jan 2021
- Developed a scalable pilot training software platform using Python, PySpark, and Hadoop to efficiently manage and process large volumes of flight simulation data, leading to a 15% reduction in training simulation run time.
- Collaborated with cross-functional teams to leverage AWS S3 for seamless data sharing, allowing training instructors to update pilot performance data remotely implementing Spark-SQL queries.
- Conducted software testing for the simulator test bed module including unit, integration tests and code coverage checks to ensure software quality, collaborating with teams in France and UK to achieve code coverage of 98%.
Systems Engineer, Radiant Coral Digital Technologies Pvt. Ltd. May 2018 – Feb 2019
- Developed production-ready autopilot software (C++) for Unmanned Aerial Vehicles (UAVs) used in surveillance, meeting project requirements and achieving an improved accuracy of 35% and full-flight stability.
Projects
INTERVIEW ASSISTANT APP DEVELOPMENT (March’ 24)
Engineered a Streamlit-based full-stack interview preparation application analyzing resume and job description to deliver customized mock interviews & feedback, enhancing user experience through a seamless real-world simulation. Incorporated OpenAI’s Whisper for transcription fidelity and Anthropic’s Claude Opus LLM model using API calls for adaptive questioning & scoring, underpinned by strategic prompt engineering for personal development insights. [Demo][Github]
PORTFOLIO DEFAULT RATE ANALYSIS (Nov’ 23)
Directed data-centric initiative to forecast one-year default probability on real-world dataset of 5M credit records. Spearheaded the project's entire lifecycle, integrating financial ratios for data imputation, cleaning, sanity checks and classification, ensuring comprehensive analysis accommodating business practices and financial theory. Developed and optimized explainable machine learning models via walk-forward analysis, achieving 15% improvement in AUC Scores over traditional benchmarks (XGBoost: 0.857, Logit: 0.827, Altman Z-Score: 0.748). [Github]
AMERICAN SIGN LANGUAGE RECOGNITION (May’ 23)
Innovated transformer-based model for Kaggle Competition, attaining remarkable 99% accuracy & swift 22 FPS performance on NYU’s HPC cluster. Streamlined model to compact 7MB using TFLite, amplifying efficiency for real-time ASL recognition & enhancing learning interactions. [Github]
MUSIC RECOMMENDATION SYSTEM (Apr’ 23)
Designed ALS-Latent factor model on Google Dataproc, leveraging 180M+ interactions of 10k users & 18M+ tracks from ListenBrainz. Outperformed popularity-baseline model with 10X boost in Average Precision. Implemented LightFM variant cutting training time by 8X with 300 latent factors. [Github]
FORECASTING STOCK RETURNS (Mar’ 23)
Crafted statistical models for portfolio allocation employing financial-model-free Reinforcement Learning frameworks on Poloniex dataset, attaining 4-fold returns within 50 days. [Github]
JUST BREATHE (Dec’ 22)
Conceptualized a microcontroller device with stress sensors using real-time embedded signal processing in C++ to identify Sudden Infant Death Syndrome (SIDS) in infants under 1 year with 97% precision. [Github]
FACTORY PLANT SIMULATION (Apr’ 22)
Crafted a sophisticated multi-threaded C++ program for a dynamic manufacturing plant scenario, ensuring efficient production and assembly workflows under strict time constraints. Incorporated advanced thread synchronization and resource allocation techniques, achieving sub-100ms runtime and effectively handling buffer capacities and wait times, showcasing enhanced operational efficiency.[Github]
STORING LOGIC CIRCUITS (Dec’ 21)
Formulated C++ code using permutation indexing in hash functions, ensuring unique and collision-free truth tables for combinational logic circuits with 94.3% accuracy rate. [Github]
Academic Honors/Achievements
Awarded the Myron M. Rosenthal Award for Best MS Academic Achievement in Computer Engineering at NYU.
Authored 3 journal and 2 conference papers (21 citations) in top peer-reviewed platforms [Google Scholar].
Secured 16th Rank in Regional Math Olympiad 2013 and qualified for Indian National Maths Olympiad.
Recipient of the University Fellowship at Syracuse University (SU) awarded to top 10 meritorious students at SU.
Served as Course Assistant in Machine Learning (ECE-GY 6153), Probability & Stochastic Processes (ECE-GY 6193) and Deep Learning (ECE-GY 7153) courses, providing support & clarified complex concepts for 200+ students at NYU through recitations and Q&A sessions.
Won silver medal in Chess at the General Championship, Sports, and Games at IIT Kharagpur.
Amongst top 0.2 percentile from over 1.5 million students in IIT-JEE (engineering exam) in 2013.