I'm working on algorithmic trading workflows using broker APIs, and I wanted to understand how other developers validate their strategies before connecting them to live order execution.
Before deploying any strategy, I usually try to answer a few questions:
How much historical data do you use for testing? Do you perform walk-forward validation? How do you measure maximum drawdown? Which performance metrics matter most to you? Do you test in paper trading before enabling live execution?
From my experience, writing the API integration is often the easier part. The bigger challenge is ensuring that the trading logic is robust enough before it starts placing real orders.
I'm interested in learning how other Kite Connect developers approach this process.
A few questions for the community:
How long do you backtest a strategy before going live? Do you rely only on historical results, or do you also use paper trading? Which metrics are mandatory before deployment? How do you monitor strategy performance after it starts running?
I'd appreciate hearing about your workflow, tools, and any lessons you've learned from deploying automated strategies.
