DataBridge AI — End-to-End Analytics Platform (Web + Desktop)

"Analysts often juggle separate tools for data cleaning, visualization, machine learning, and AI-assisted analysis — switching between spreadsheets, notebooks, and dashboards for a single workflow. I wanted to prove a single platform could handle the entire pipeline, from raw file to trained model, without leaving one interface."

"I built DataBridge AI, a Streamlit-based platform covering the full data lifecycle: import from CSV, Excel, JSON, Parquet, SQLite, or a live database connection, with automatic header detection and cleanup on upload. From there, users move through AI-assisted column mapping, a 0–100 data quality engine, outlier detection, feature engineering (encoding, scaling, binning), and a no-code ML Studio supporting classification, regression, and clustering."

"An integrated AI Assistant answers natural-language questions about the dataset, and a dynamic KPI tracker calculates target achievement automatically."

"To extend its reach beyond the browser, I packaged the platform as a native Windows desktop application using Tauri — bundling an embedded Python runtime so it runs as a standalone .exe, no separate installation required."

"The result is a single tool that works two ways: as a live web app for quick access, and as an installable desktop application for offline or client-side use. Built to demonstrate range across data engineering, machine learning, applied AI, and software packaging — not just a dashboard, but a shippable product."

Live demo: https://databridge-ai-epgqwj77nmmqg6bsseardf.streamlit.app/ Username: demo . Password: CH8yxiGCHMHs7T5

(demo mode, no API key required)