kashifa User

I've built end-to-end analytics projects that turn raw data into forecasts, risk signals, and actionable reporting.

In my supply chain risk prediction project, I engineered a spatio-temporal graph neural network combining GCN and LSTM architectures to forecast demand and flag emerging risks across network nodes. I evaluated predictive reliability with ROC-AUC and F1-Score, then deployed an automated real-time risk-reporting pipeline using FastAPI, PyTorch, and Google Colab.

I've also analyzed multi-year stock price data using time-series decomposition and developed tuned ARIMA models to identify trend, seasonality, and cyclical patterns for investment analysis.

As a Computer Science Engineering graduate, I bring Python, SQL, statistical modeling, Power BI, Tableau, and Excel to data cleaning, visualization, and report automation work.