Back to Projects

AI-Assisted Data Dashboard

AI-Assisted Stock Analysis Dashboard

An AI-assisted data analysis dashboard project that turns personal stock research questions into an interactive dashboard through requirement breakdown, modularization, and visualization iteration.

Period

Aug 2025 - Present

Tools / Tech

PythonStreamlitAKSharePlotlyPandasScikit-learn

Why I built it

Project focus: the dashboard is for observing intraday signals, order flow, large orders, fund-flow estimates, and data patterns. It is not a trading recommendation tool.

Links

What it includes

  • Shows intraday quotes, tick transactions, large-order behavior, fund-flow estimates, and price-level changes.
  • Separates observation indicators from trading recommendations and states clearly that the output is for review only.
  • Includes related AKShare API checking work for testing endpoint availability and data-report outputs.
  • Includes experimental anomaly detection, classification, and clustering modules as research examples.

What I worked on

  • Used AI assistance to split data fetching, cleaning, indicator calculation, chart display, and module structure.
  • Reviewed AI-generated code and charts manually to keep field handling, empty-data cases, and visual output understandable.
  • Kept the project framed as exploratory analysis rather than a financial product.

What I Learned

  • Learned how financial-data tools need careful wording around observation versus advice.
  • Practiced AI-assisted modularization for data dashboards.
  • Improved awareness of data-source instability, field changes, and visualization edge cases.