Financial Portfolio Analysis

I built this with Julien and Pierre-Yves during the Le Wagon Data Engineering Bootcamp. It collects market data from Yahoo Finance and economic series from FRED, then moves the data through Google Cloud Storage, BigQuery, and dbt before it reaches the Streamlit dashboard.

Overview

Airflow schedules the ingestion jobs. Raw Parquet files land in GCS, BigQuery exposes staging and mart tables, and dbt turns them into daily prices, returns, volatility, Sharpe ratios, moving averages, and drawdown metrics for roughly 460 tickers.

What I worked on

The project covers the full path from extraction to analysis: retries and validation during collection, date-partitioned storage, warehouse models, and a dashboard with market overview, ticker deep dives, benchmark comparisons, and portfolio construction.

The dashboard is available at alphaview.streamlit.app.