May 2024 – Jul 2024 · 2 mos · Bengaluru, Karnataka, India · On-site

• Objective : Graph-RAG Pipeline for In-Depth Stock Market Analysis, providing Explaination of Price Movements and Forecasting Risks and Opportunities
• Developed a Graph-RAG pipeline that boosted stock explainability by 15.78%, pinpointing reasons for price fluctuations and forecasting future risks
• Achieved a 23% performance boost over GPT-4o on the RAGAS framework, significantly enhancing accuracy and reliability in stock market analysis
• Reduced monthly costs from $22,000 to $1,810 using Graph-RAG over GPT-4o, providing real-time stock insights through a refined chatbot

Jun 2023 – Aug 2023 · 2 mos · Stanford, California, United States

• Enabled efficient healthcare planning by designing a robust framework for forecasting 6 primary health indicators in low, middle-income countries (LMICs)
• Reduced feature count by 75.5% & handled missing data by leveraging dimensionality reduction & Imputation on a dataset of 0.5 Million datapoints
• Achieved a base 63% improvement in predictions by applying PCA and XGBoost ML techniques, following experimentation with over 10 other models

Jun 2023 – Jul 2023 · 1 mo · Paris, Île-de-France, France

• Uncovered valuable insights by extracting reviews from TripAdvisor and Expedia, of over 100,000 hotels across the top 100 US cities, over a 2-year period
• Investigated chatGPT's impact on fake reviews

Jun 2023 – Jul 2023 · 1 mo · United States

• VC Due Diligence and deal sourcing: Unearthing potential of Pre-Seed and Seed Beauty and Wellness startups for strategic Investment consideration
• Performed due diligence on pre-seed Beauty and Wellness startups and Prepared a well crafted Investment Memo,

May 2023 – Jul 2023 · 2 mos

• Conducted comprehensive scraping and processing of over 50,000+ Airbnb images and price details of 10,000+ listings across 20+ prominent US cities
• Evaluated of 11+ key image features, encompassing colorfulness, sharpness, objects, saturation, color balance, and more, for a comprehensive analysis
• Correlated image features with the listing prices, achieving a 72% accuracy by applying machine learning techniques like Linear Regression & XGBoost

Mar 2023 – Aug 2023 · 5 mos · Cambridge, England, United Kingdom

• Empowering Micro-entrepreneurs through effective mentorship and data-driven progress monitoring to foster Sustainable growth
• Employed ML algorithms to evaluate 500+ store images, accurately detecting and monitoring over 150+ crucial elements for overall progress assessment

Jul 2022 – Nov 2022 · 4 mos · Pennsylvania, United States

• Enhancing Safety in US National Parks: Leveraging Deep Learning for Human-Wildlife Interaction Monitoring and Analysis
• Analyzed over 5000+ images to monitor human-wildlife interactions in US National Parks, providing actionable insights and preventive measures for safety
• Reduced manual efforts by over 150+ hours, by everaging Deep Learning techniques to precisely identify abnormal human-wildlife interactions

Apr 2022 – Jul 2022 · 3 mos

• Developed and trained a Machine Learning algorithm utilizing UNet architecture for Image Segmentation of satellite images
• Reduced computation time by 55% through Gaussian Blur to blur the backgrounds thereby highlighting critical features for enhanced visual analysis