Jul 2023 – Oct 2023 · 3 mos · Bengaluru, Karnataka, India · On-site

• Winner of Shell.ai General Edition Hackathon, competing against 5700+ Data Scientists in Shell's flagship event Changemakers of Tomorrow for sustainable and affordable energy
• Integrated publicly available crop production, population census and land cover reports at coordinate level along with engineered spatial and administrative features of Gujarat state
• Deployed an ensemble of XGBoost and Light GBM Machine Learning models to forecast biomass production on 2416 different geographic locations at an MAE of 30.23
• Modified K-Means++ smart initializer and K-Means clustering algorithm from scratch to identify optimal depot and refinery locations while minimizing operational and transportation costs

Nov 2022 – Mar 2023 · 4 mos · Bengaluru, Karnataka, India · Remote

• Clinched the 7th rank among a pool of 2000+ professional data scientists across the globe in the World's biggest Industrial AI challenge
• Enhanced model performance and data variability through artificially synthesized images of multiple opaque and see-through cropped objects
• Brainstormed Intelligent Video Analytics to build a real-time inventory system and identify the status of industrial assets like KLT boxes, racks and placeholders
• Quantized model weights of yolov7x object detection model to FP16 in TensorRT format for fast video inferencing, achieving 190 frames per second on 2-minute real-time rendered video
• Deployed models and NvDCF multi-object tracker in dual integrated Deepstream SDK pipelines to achieve an mAP score of 90% in a live demonstration to the key stakeholders

Jan 2022 – Aug 2022 · 7 mos · Greater Bengaluru Area

• Leadership and Process Improvements
• Spearheaded the automation of traditional practices and SAS scripts that led to a
• significant reduction of manual work across department
• Distributed data processing into large datasets via SAS parallel sessions to
• improve its processing speed by 75%
• Collaborated with multiple teams across the business to develop and implement Delinquency Framework for the different US unsecured portfolios
• Identified and debugged the potential issues in existing workflow across multiple
• projects to enhance process standardization
• Key Achievements
• Received the Citi Gratitude Award for distinguished performance and multi-tasking
• across varied projects under strict timelines

Aug 2020 – Jan 2022 · 1 yr 5 mos · Greater Bengaluru Area

• Model Development, Governance, and Analytics
• Responsible for regulatory modeling of the US unsecured branded cards portfolio
• to meet the requirements of key stakeholders in revenue and capital planning
• Administered end-to-end pipelines for the development, testing, and implementation of
• Liquidation Rate Model for CECL purposes, first of its kind in the industry
• Conceptualised factor-based framework for Delinquent Balances and Units as a
• the function of Gross Credit Losses for portfolios of over 30 billion accounts
• Revalidated State Transition Matrix-based CCAR Challenger Model to assess the
• impact of COVID-19 in compliance with Model Risk Management guidelines

Oct 2019 – Oct 2019 · 0 mo · Gurgaon, India

• Ranked 3rd among over 2000 participants from all IITs in American Express flagship hackathon
• Performed feature engineering and missing value imputation using tuned machine learning models
• Trained XGB Classifier and accurately predicted the credit line of customers based on past records

Sep 2019 – Sep 2019 · 0 mo · Gurgaon, India

• Ranked 4 among over 1500 students from the top IITs in Gartner Hackelite 2019
• Implemented feature engineering which better represented the problem
• Trained XGB model to achieve an F1 score of 90.4% in predicting Gartner client retention

Aug 2019 – Aug 2019 · 0 mo · Chennai Area, India

• Among the top 25 nation-wide finalists in the flagship event of IIT-M Alumni Association in association with Wells-Fargo
• Built Prophet model and forecasted traffic volume for next year with an RMSLE of 0.12%

Jul 2019 – Jun 2020 · 11 mos · Chennai, Tamil Nadu, India

• Devised a framework of Artificial Neural Networks to optimize Jacket structure
• dimensions using various weather and other parametric simulation data
• Simplified the process of Jacket structure design to significantly reduce the trial and
• error iterations of the conventional technique
• Key Achievements
• Awarded the Prof C.S. Krishnamoorthy Endowment Prize for the best research
• project in Genetic algorithms and Evolutionary Computation at the Institute level

Jul 2019 – Jul 2019 · 0 mo · Bengaluru Area, India

• All India rank 288 among 8000+ participants including students from top IITs
• Tuned hyperparameters of RF3 and XGB models using Random Search and Grid Search CV
• Achieved an accuracy of 85.6% in predicting the probability of Cristiano Ronaldo scoring a goal in a football match

May 2019 – Jul 2019 · 2 mos · Gurgaon

• Performed predictive analysis on B2B sales and marketing data to identify promising
• deals
• Implemented log transformation and bucketed column feature engineering along
• with z-score analysis for the outlier removal of the data
• Built a Random forest model on 6 fiscal quarters to predict the deals won for the future quarter and improved F1 score by 4.49%

May 2018 – Jul 2018 · 2 mos · Greater Philadelphia Area

• Sentiment Analysis on Cyber Security dataset
• Analysed user reviews from cybersecurity forums using lexical resource
• SentiWordNet produced by 8 ternary classifiers
• Established a rule-based approach framework to calculate cyber-attack sentiment
• score and correlated with financial impact of cyber attack
• Extracted Parts of Speech using linguistic feature selection to calculate score;
• outperformed IBM Watson API results
• Nominated for Best Student Research Paper Award at the 2019 Northeast
• Decision Sciences Annual conference
• Sentiment Analysis on Wine dataset
• Engineered a recommendation system to match optimal wines for users using
• opinion mining on more than 3L wine reviews
• Recommended wine variety using the wine descriptions and prices by applying count
• vectoriser in Extra Tree Classifier
• Generated heat map using K-Means Clustering to map out varieties of wines and
• their counts in each of the cluster

Jun 2017 – Jul 2017 · 1 mo · Lucknow Area, India

• Analyzed 3 Harvard Business Publishing Cases
• Established relations between various factors and the dependent variable
• Visualized the given sample data using Frequency Tables, Correlations, T-Tests
• Hotel Management Project
• Explored data of 42 Indian cities and performed price analysis of hotel rooms using correlations and manager intuition
• Developed a Linear Regression Model based on star rating, swimming pool and tourist destination to predict rent for any room

May 2017 – Jun 2017 · 1 mo · Cochin Area, India

• Studied the construction of India’s first indigenous aircraft carrier INS Vikrant and commercial, patrol and survey vessels
• Observed material flow and workflow through detailed studies in project planning and quality control