Yash Raj
Mathematics & Scientific Computing student with a focus on data science and backend engineering. Building data pipelines, designing REST APIs, and applying statistical methods to extract insight from real-world datasets.
About
My mathematical background shapes how I approach model design, data quality, and performance trade-offs. I work on building data pipelines, designing REST APIs, and applying statistical methods to extract insight from real-world datasets.
Currently deepening my work in predictive analytics and machine learning — moving from descriptive analysis toward building systems that forecast and classify.
Technical skills
Projects
Automated Wealth & Portfolio Optimization API
View repo →An asynchronous backend microservice that ingests real-time EOD price data and US Treasury rates from the Financial Modeling Prep API and computes the mathematically optimal capital allocation across a multi-asset portfolio. A 24-hour background sync worker decouples data ingestion from the API layer, keeping response times low regardless of external API latency.
Dynamic Pricing Optimization Engine
View repo →A high-performance microservice that computes the mathematically optimal price for products to maximize profit. Uses constrained minimization to locate the exact peak of a profit function by balancing base costs, dynamic demand multipliers, and competitor pricing. PyArrow loads a compressed Parquet database directly into RAM on boot, enabling Pandas Boolean indexing across 100,000+ synthetic products in milliseconds.
E-Commerce Customer RFM Segmentation API
View repo →An end-to-end data science pipeline and REST API for customer segmentation. Processes 100,000+ real-world e-commerce transactions to compute Recency, Frequency, and Monetary scores and classify customers into actionable business segments in real time.
Currently learning
SQL
Query optimization and working with large-scale structured datasets.
Scikit-learn
Supervised and unsupervised learning pipelines.
Machine Learning Fundamentals
Regression, classification, and clustering with emphasis on the underlying mathematics.
Pipeline Optimization
Applying scientific computing principles to improve data processing efficiency.
Contact