Workflow Overview

The closed loop of configure, operate, and adjust, with key points for each phase.

The core workflow for using the system is: Setup (exchange, AI, notifications) → Operate (create strategy → run backtest → paper trading → live trading) → Adjust (observe, optimize, with the AI Agent's help if you like). The sections below cover the key points and the matching docs for each phase.

Flow diagram

Configure
Exchange / AI / Notifications
Operate
Strategy → Backtest → Paper → Live
Adjust
Monitor / Optimize / AI
Create StrategyBacktestPaper TradingLive Trading
After adjusting, loop back into the operate phase
  • Operate phase: Create Strategy → Run Backtest (per-pair execution config) → Paper Trading → Live Trading; after the Adjust phase you can come back here and iterate on the strategy or execution config.
  • Adjust phase: after a backtest, review K-line charts and individual entries / exits and re-run backtests with tweaked parameters; while running, observe and optimize based on trade records and performance, and let the AI Agent analyze or adjust the strategy / execution config.

Operate phase

Create a strategy

In the Strategy Library, pick a template or configure buy / long & sell / short conditions from scratch, and set the indicator parameters used in those conditions. See Strategy Basics, Condition Combinations, Technical Indicators, Create Strategy.

Run a backtest

Create a backtest task (pick the exchange, spot / futures, time range), add pairs to the task, and for each pair configure the K-line timeframe, initial capital, the strategy to use, and the execution config; after launch, review return, win rate, max drawdown, and other metrics before considering paper or live. See Run Backtest, Create Backtest, Result Analysis.

Paper Trading

Create a trading bot, pick Paper Trading, and run with virtual capital under live market conditions for a while before going live. See Paper Trading.

Live Trading

Once paper passes, switch the trading bot to Live Trading and use real capital; the exchange and API need to be configured. See Live Trading (workflow), Live Trading, Risk Management.

Adjust phase

Adjustment isn't limited to runtime—you can start as soon as a backtest finishes:

  • After a backtest: review K-line charts and each entry / exit, tweak strategy parameters or per-pair execution config and re-run backtests until you're satisfied, then go to paper or live.
  • While running: review Trade Statistics and aggregate stats, combined with Monitor, to diagnose performance; adjust parameters or logic and—if needed—re-validate via backtest before updating live.
  • At any time, ask the AI Agent to analyze performance, interpret data, suggest adjustments, or directly modify the strategy / execution config.

Setup phase (before use)

Before you operate and trade, complete: Exchange connection (required for live), AI configuration, Notification configuration. You can also do the whole setup through the AI Agent by chatting with it (e.g. adding exchanges, APIs, notification channels), and it will walk you through step by step. See the first step of the usage flow in the Introduction.

Recommended practice

  • Paper before live: once the backtest passes, run Paper Trading for a while, then switch to Live Trading.
  • Start small: validate live with small capital first, then scale up.
  • Risk first: when creating a backtest or bot, set sensible per-pair execution config (stops, position, drawdown)—see Execution Config, Risk Management.

Next steps