AI Crypto Trading Web App

I’m building a fully-featured, AI-assisted crypto trading and back-testing application that runs flawlessly on both desktop and mobile screens. The stack is set: a FastAPI (Python) backend for speed, a React front end for a crisp user experience, and an embedded TradingView widget to power the charting.

Here is what I expect when the job is complete:

• A dark-themed, mobile-responsive interface that loads BTCUSDT, ETHUSDT, SOLUSDT and any custom symbol I enter.
• The EMA 20/50 crossover strategy pre-wired, with RSI, MACD and Volume overlays and clear Buy/Sell signals plotted directly on the chart.
• Risk/Reward and Position Size calculators that sit alongside the chart so a user can size trades instantly.
• One-click AI screenshots of the current analysis saved to a Trade Journal, accompanied by an automatically generated equity curve, total return, win rate, Sharpe ratio and maximum drawdown. Historical back-testing must stretch a full 10 years and complete quickly.
• A secure login flow with separate Admin and User roles. An Admin dashboard should let me manage users, symbols and any system parameters.
• The code needs to be clean, documented, and production-ready, with the source delivered in a private repository.

Acceptance criteria
1. All calculations match TradingView results to within 0.1 %.
2. Back-tests on BTCUSDT over the full data set complete in under 20 seconds.
3. Lighthouse mobile performance score ≥ 90 on a mid-range Android device.
4. Complete README covering local setup, deployment and extension points.

If this sounds like a challenge you’re ready to tackle, let’s talk timelines and milestones—then dive straight into the code.