Backend API

This section outlines the core APIs exposed by the TradeSim backend for trade simulation, real-time market data, and machine learning-based prediction services.

WebSocket API

The backend streams real-time data via WebSocket for high-frequency updates.

Purpose	Endpoint	Description
Prediction Metrics	/ws/prediction	Streams midpoint price predictions and metadata computed by the ML model.
Orderbook Data	/ws/okx	Streams live orderbook updates from the OKX exchange for BTC-USDT_SWAP.

WebSocket Message Format

• All WebSocket messages are JSON-encoded using orjson for high-speed serialization.
• Clients must maintain persistent connections to receive continuous updates.
• Message content depends on the type of stream:

Example: Orderbook Stream

{
    "timestamp": "2025-05-24T10:22:29Z",
    "exchange": "OKX",
    "symbol": "BTC-USDT-SWAP",
    "asks": [
        [
            "108225.1",
            "832.32"
        ],
        [
            "108225.2",
            "0.66"
        ],
    ],
    "bids":[
        [
            "108225",
            "348.79"
        ],
        [
            "108224.7",
            "0.11"
        ],
    ]

Example: Prediction Stream

{
    "prediction": [
        108950.20113248515
    ]
}

WebSocket data is pushed in real-time with minimal latency, thanks to FastAPI's async capabilities.

REST API

Simulate Trade

Purpose	Endpoint	Method	Description
Simulate a Trade	/simulate/trade	POST	Simulates a market trade and returns execution metrics like slippage, fill, latency, etc.

Request

{
  "asset": "BTC-USDT_SWAP",
  "exchange": "OKX",
  "orderType": "market",
  "quantityUSD": 1000
}

Response

{
    "timestamp": "2025-05-24T17:18:53Z",
    "exchange": "OKX",
    "asset": "BTC-USDT-SWAP",
    "order_type": "market",
    "qty_usd": 5000000.0,
    "filled_qty": 45.91403658,
    "fee_usd": 2000.0,
    "slippage": 0.0,
    "average_price": 108855.6,
    "total_spent": 5000000.0,
    "latency": "0.24ms",
    "status": "success"
}