Yash Sethi
Nice to meet you! 👋
I'm a Master's student in Management Analytics at McGill University. Passionate about using data science to solve real-world problems and uncover insights through data.
Dive into my portfolio and blog to explore how I’m harnessing data to shape the decision-making of businesses.
01. About Me
Hi! I'm Yash Sethi, a 27-year-old from India with a passion for data, problem-solving, and continuous learning. My journey into the world of analytics started with a love for mathematics, which I pursued through both my Bachelor’s and Master’s degrees.
Over the past four years, I’ve had the privilege of working at HSBC, where I transitioned from marketing analytics to fraud mitigation. From designing hyper-personalized product recommendation models that saved millions in marketing costs to developing fraud detection systems that prevented significant losses, my work has been all about turning data into actionable insights.
Currently, I’m expanding my expertise by pursuing a Master’s in Management Analytics at McGill University. Alongside my studies, I’m working as a Data Analyst at ALDO and leading a project as a Data Scientist with BNP Paribas, further sharpening my skills in real-time analytics, forecasting, and predictive modeling.
02. Where I’ve Worked
Data Scientist @ Assets Liability Management & Treasury (ALMT) Team
Montreal, Canada
Sep 2024 – Present
- Planned and implemented a framework to automate data collection using Bloomberg plugins for macroeconomic data, Python for refreshing Excel and feature aggregation, resulting in a ready-to-use dataset for forecasting models, saving one FTE monthly.
- Modelled and compared ARIMA (non-stationary time series), Bayesian Structural Time Series (probabilistic), and Time-Series GAN (TSGAN) models to forecast the spread between 10-year and 2-year Treasury yields. The Bayesian Structural Time Series model outperformed others, achieving a monthly RMSE of 0.12 and a direction accuracy of 62%.
03. Notable Projects
All-Weather Fundamental Surprise Enhanced Momentum Strategy
A cutting-edge investment strategy developed during the McGill-FIAM Asset Management Hackathon 2024, where our team secured 4th place. This strategy enhances traditional momentum approaches by addressing their weaknesses at market turning points, especially in mid and small-cap universes, using fundamental surprises and macroeconomic regime detection.
Toronto Fire Incidents Analysis and Prediction
A comprehensive analysis of Toronto fire incidents using data from the city's Open Data portal. This project explores historical fire incident trends, performs in-depth data cleaning and visualization, and applies machine learning models to predict the estimated dollar loss caused by fire incidents at various locations and due to different causes.
Automobile Price Predictor
A project focused on predicting automobile prices using exploratory data analysis (EDA) and machine learning techniques in R. This project leverages statistical methods to uncover patterns in data and builds robust predictive models to estimate car prices accurately. The analysis includes data preprocessing, feature engineering, and comparisons of different ML algorithms.
StormShield: Optimization for Emergency Evacuation
Designed an optimization model to enhance emergency evacuation planning for hospitals in hurricane-prone regions. The model leverages data on hospital capacity, patient criticality, and resource constraints to allocate ambulances, air ambulances, and buses efficiently. The goal was to minimize transportation costs, evacuation time, and carbon emissions while ensuring patient safety during hurricanes.
Optimal Parameters for Type-II Censoring Experiments
A research project conducted during my masters at IIT Kharagpur, focusing on designing Type-II censoring experiments to compute percentile points for the Kolmogorov-Smirnov statistic. The project aimed to optimize experiment parameters for accurate estimation of data distributions under Type-II censoring.
04. What I’ve Studied
Master of Management, Business Analytics @ McGill University
Montreal, QC, Canada
Jul 2024 – Present
Courses Taken:
- MGSC 660 - Mathematical and Statistical Foundations for Analytics
- MGSC 661 - Multivariate Statistics
- MGSC 662 - Decision Analytics
- INSY 695 - Machine Learning and Deep Learning for Enterprise
- MGSC 673 - Intro to AI & Deep Learning
- MGSC 695 - Decision Analytics and Modelling for Operations
- INSY 660 - Coding Foundations for Analysis
- INSY 661 - Database & Distributed Systems for Analytics
- INSY 662 - Data Mining and Visualization
- INSY 669 - Text Analytics
- INSY 670 - Social Media Analytics
- ORGB 660 - Managing Data Analytics System
- ORGB 661 - Ethical Leadership and Leading Change
05. Read My Blogs
Interested in learning more about my thoughts and experiences? Read more to explore my blog posts and insights on the Blog Page ↗.
Get In Touch
I am always open to opportunities!
Please feel free to drop a quick hello or ask anything, and I will get back to you soon 😊.