Data Science Graduate Student with 3+ years of industry experience & seeking full time opportunities.
Data Analyst Intern, The Prefect Child LLC, Brooklyn NY
Jun 2022-Aug 2022
- Built a summary dashboard, capturing appropriate KPIs and statistics.
- Created a database for addresses & geo coordinates using google maps API, and visualized on a map.
Associate Data Scientist, Merilytics, India
Apr 2020-Aug 2021
Built an Automated Valuation Model (AVM) using TensorFlow-based custom nearest-neighbor architecture to identify comparable properties for 4 major property types.
- Delivered ~$25 million annual savings by reducing ~2700 man hours weekly by automating property valuation.
- Collaborated with client on biweekly calls to deploy the AVM in the client's environment by establishing an end-to-end data pipeline, & refactoring the code using PEP guidelines.
- Built a Django Website and hosted the AVM on it to input with ease, and display the results (comparables) on a map.
Developed a Demand forecast model with 1800 SKUs for a major European EV supplier
- Leveraged Time Series forecasting packages such as fb-prophet, neural prophet, & seq2seq neural nets.
- Reduced time taken for forecasting from 1 week to 4 hours with the automated pipeline.
Created a heuristic based driver scheduling algorithm to incorporate driving, terminal, & regulatory constraints.
- Automated driver scheduling, reducing the time taken from multiple days to 15 minutes.
- Produced a simulation of bills movement across 34 terminals for a long-haul trucking client.
Senior Data Science Analyst, Merilytics, India
Feb 2019-Apr 2020
- Built Sales Forecast Model using Keras for creating promotion strategy for an American online clothing chain.
- Developed an end-to-end data pipeline on Azure utilizing Azure functions to automate ETL.
- Utilized KNN for determining comparable real estate properties, incorporating weights for different features.
Research Intern, Vidooly, India
Dec 2018-Jan 2019
- Developed several Keras models like Xception and ResNet for classifying YouTube Thumbnails.
- Established the data pipeline for extracting thumbnails using YouTube API.
Python Developer Internship, Modestreet, India
Sep 2017-Dec 2017
- Created a web app using Django & Three.js to make 3D human models with specific dimensions, to be used as a virtual mannequin for an online clothing store.
Leadership Experience
President, UB Data Analytics Club, Buffalo NY
Apr 2022-Dec 2022
- Spearhead events for weekly meetings regarding resumes, cover letters, and LinkedIn workshops, presenting research for 200+ students.
Co-Founder & Content Creator, Synergy Learn, India
Jan 2018-Sep 2018
- Animated & Produced lectures for YouTube & garnered 10,000 hours of watch time.
- Managed a graphic designer, company website, and social media to promote business.
Master of Science, Data Science, University at Buffalo, SUNY
Sep 2021-Jan 2023
Relevant course work: Statistical Data Mining, Data Model & Query Language, Probability Theory, Predictive Analytics, Machine Learning, Reinforcement Learning.
GPA: 3.8
Bachelor of Technology, IIT Roorkee, Industrial Engineering
May 2013-May 2017
- Languages : Python, JavaScript, MATLAB, R, SAS
- ML Tools : Classification, Regression, Clustering, Tree Based Algorithms, Bagging & Boosting.
Libraries: TensorFlow, Keras, PyTorch, and scikit-learn. - Deep Learning & NLP : LSTM, Transformers, CNN, YOLO, Word2Vec, TF-IDF, Topic Modelling, spaCy, NER
- Data Management : SQLite, MySQL, pandas, Numpy, Excel
- Tools & Data analysis : Version Control (git), Power BI, Tableau, Django, JavaScript, Time Series Forecasting, Data Visualization, Machine Learning Interpretability
New Architecture : Created a new deep learning architecture based nearest-neighbors.
Reinforcement Learning : Created a scalable multi agent environment for Ludo from scratch, used Tabular methods with the downscaled version of Ludo & actor critic for full-scale version, and Designed the reward structures to accelerate learning.
Dashboard : Created an e-commerce platform from scratch with a summary dashboard, orders page, and a linked PostgreSQL database.
Time Series Forecasting : Time Series Analysis, customer segmentation, and interactive dashboard for 100k orders from an e-commerce platform.
Resume Optimization :
Managed 6 profile versions on a single spreadsheet to keep changes in sync & avoid repetitive editing.
Utilized Python, HTML, & CSS for formatting & rendering, and Excel for handling data to create this version of my resume.
Utilized Python, HTML, & CSS for formatting & rendering, and Excel for handling data to create this version of my resume.
Ongoing
8 Ball Pool : Predicted & visualized ball trajectories using OpenCV, & made preemptive optimal decisions.
Smartphone Price Prediction : Mined data from GSMArena and performed feature engineering by mapping Centurian Mark Score using fuzzy logic
Wildfires Analysis : Visualized the clusters of different wildfires using geopandas, and predicted the Arson wildfire by performing EDA & appropriate feature engineering.
Poultry Price Forecast : Scraped 2 years of data using Selenium and built a Time Series forecast model.
Probability Project : Found expected values using analytical approach as well as simulation.
Clustered grocery items using kmeans for optimal positioning & proximity of similar items.
Sentiment Analysis : Used Glove Word Embedding (NLP) for analyzing sentiments from twitter.
Convolutional Neural Network : Used Keras to create a binary classifier using Keras.
Document blender website : Documents with same structure are maintained in different google doc files, and sliders for each section of the document are used to blend them.
Used spreadsheet & javascript to create a table in backend, where each row forms a section of documents.
Used spreadsheet & javascript to create a table in backend, where each row forms a section of documents.
Achieved 99.76 Percentile in the Joint Entrance Examination (All India exam with 1.4 million candidates).
Achieved 98.98 Percentile in eLitmus pH Test for Problem Solving.