SYSTEM 01 // PROJECTRole: FEATURED PROJECT
GNN + LLM Integration for Stock Portfolio Optimization
Designed and implemented a GraphSAGE-based GNN using PyTorch to model relationships between Nifty-50 stocks using financial and textual features. Integrated FinBERT for sentiment extraction and Mistral-7B for reasoning-based portfolio scoring. Built end-to-end ML pipeline including data preprocessing, feature engineering, model training with gradient optimization, and backtesting. Documented methodology as a formal M.Sc. thesis and IEEE/APA research article.
TechnologyGraphSAGE
StackGNN
PlatformPyTorch
OutcomeAchieved 28.4% portfolio return vs. 19.2% baselineSelf-reported
Metric 02—Pending
Metric 03—Pending
View Project
// INTERVIEW ENGINE