🚀 LazyTrader AI

LazyTrader AI is an autonomous AI-powered crypto paper trading bot. It watches live Bitcoin and Ethereum prices, asks Google Gemini what to do, executes simulated trades, and tracks portfolio performance — all from a single interactive web dashboard.

One magic button. Full AI-driven trading session. No human judgment required.

---

Features

• ⚡ Magic Auto-Trade — one click triggers 3 consecutive AI-driven trading rounds with live animated updates
• 🔬 AI Backtest — sends 30 days of real BTC + ETH price history to Gemini in a single prompt; simulates daily trades and compares the AI strategy against simple Buy & Hold
• 📈 Portfolio Chart — live area chart of portfolio value growing over time
• 💼 Multi-asset — trades both Bitcoin (BTC) and Ethereum (ETH)
• 📋 Trade History — full log of every decision with price, amount, and timestamp
• 🔄 Reset — wipe the portfolio back to $10,000 for a clean demo

---

How it works

CoinGecko API  ──►  Live BTC + ETH prices
                         │
                         ▼
              Google Gemini (LLM)  ──►  ACTION: BUY_BTC | SELL_ETH | HOLD
                                              │
                                              ▼
                                   Paper trade executed
                                   wallet.json updated
                                              │
                                              ▼
                                   Streamlit dashboard refreshes

1. Market Data — fetches real-time BTC and ETH prices (24h change) from the free CoinGecko REST API.
2. AI Analysis — builds a structured prompt with current prices and portfolio state, sends it to Gemini, and parses a structured response (ACTION / CONFIDENCE / REASON).
3. Paper Trading — executes the trade on a simulated $10,000 portfolio stored in wallet.json. No real money involved.
4. Backtest — pulls 30 days of daily OHLC data, packs it all into one Gemini prompt, gets a decision per day, simulates the full sequence, and plots AI Strategy vs Buy & Hold.

---

Tech stack

Layer	Technology	Why
UI / App	Streamlit	Instant Python web apps with zero frontend code
AI / LLM	Google Gemini (gemini-2.5-flash)	Fast, cheap, strong reasoning for structured decisions
Market Data	CoinGecko API	Free, no auth required, real-time crypto prices
Data wrangling	pandas	DataFrame manipulation for charts and backtest simulation
Storage	JSON (wallet.json)	Simple, portable, human-readable portfolio persistence
HTTP	Python urllib (stdlib)	No extra dependency for API calls
Config	.env file	Keeps the API key out of source code

---

Requirements

• Python 3.8+
• A free Google Gemini API key

---

Setup

1. Create and activate a virtual environment:

   python3 -m venv .venv
   source .venv/bin/activate  # Windows: .venv\Scripts\activate

2. Install dependencies:

   pip install -r requirements.txt

3. Add your Gemini API key — create a .env file:

   GEMINI_API_KEY=your_actual_api_key_here

---

Usage

streamlit run main.py

The app opens at http://localhost:8501.

Demo flow:

1. Click 🚀 Run AI Backtest — watch Gemini analyze 30 days of real data and compare vs Buy & Hold
2. Click ⚡ MAGIC AUTO-TRADE — 3 live AI trading rounds with real-time updates and balloons on profit
3. Watch the portfolio chart grow

---

Project structure

.
├── main.py           # All app logic and UI
├── wallet.json       # Live portfolio state (auto-created)
├── requirements.txt  # Python dependencies
└── .env              # API key (not committed)