Upmanyu Singh Email & Phone Number

Columbia, Maryland, United States

Washington, District Of Columbia, United States

1. Performed data mining & data analysis on 75+ undergraduate Zoom transcript to extract actionable insights, utilized advanced statistics to derive insights into students’ comprehension of course material2. Developed unique blend of advanced analytics and computer science algorithms (topic modelling, summarizer and clustering) to identify key themes in Zoom transcripts with 90% accuracy in assessing course comprehension3. Built vendor evaluation tool using AHP/BWM decision analysis in Excel/VBA automating scoring matrix for 15 criteria; boosted confidence in decision making

1. Directed 4-person cross-functional team in deploying cloud data solutions with Snowflake for 2TB+ database migration, optimizing data architecture and enhancing business strategy, demonstrating project management and team collaboration2. Improved query efficiency using Python, cutting data governance & transformation time from 23 hours to 10 minutes3. Overhauled ETL pipelines employing Pandas comprehensive real estate data architecture, escalating processing efficiency by 62% with near-perfect accuracy, directly impacting key performance indicators

1. Engineered segment tree algorithm in Python, leveraging business intelligence & econometrics to optimise $40M in credit fund allocation, accurately identifying high-value customer segments to refine go-to-market strategies & product management2. Translated client analytics data into business insights for optimized credit card offers, aligning predictive models with business objectives and presenting in executive reviews; showcasing documentation and communication skills3. Drove yearly run-rate revenue improvement from $8M to $20M+ by boosting eligibility and offer effectiveness for co-branded card credit line increases; showcasing skills in data analytics

Mumbai, Maharashtra, India

1. Saved 60+ hours/month by building end-to-end data driven NLP pipeline using Python for rapid sentiment analysis on news data, providing domain-specific insights to accelerate business performance2. Co-developed credit risk classification model using neural networks, achieving a 90% recall score in evaluating daily bankruptcy probabilities from 10K SEC filings, reflecting deep financial analysis and market research acumen

Mumbai, Maharashtra, India

1. Developed a predictive model using Machine Learning that standardized the security information coming from different sources andextracting information from the string, saving a minimum of 45 days of manual work, and a decrease in manpower from 10 people to 3 people (Just for Quality Check).2. Created portfolio management application combining data models like Black-Litterman and CPPI; integrated SQL database into stack migration from Node.js to Django REST, reducing latency by 280ms

Bengaluru Area, India

1. Developed maternal health app’s question routing system utilizing multi-label classification models and data modeling techniques, efficiently handling 90,000 daily patient queries with precise specialist matching2. Engineered label prediction Flask API for web integration, leveraging business intelligence and data structures to enhance self-service capabilities, significantly benefiting over 50,000 users with improved data reporting and product features3. Enhanced user experience by automating forum post labeling and redirection, resulting in 30% increase in app usage time

Master Of Science - Ms, Data Science, 3.8

Bachelor'S Degree, Electrical Engineering