I’ve been algo trading for about 10 years now – from some super basic setups initially (single computer at home) to cloud based strategies to hybrid models where I backtest at home but deploy the code in the cloud to using AWS lambda. Over this time this has resulted in spaghetti code (having over 110 strategies with only a fraction of them delivering actual results didn’t help either) little to no documentation on the whole thing and a setup that feels like it cannot easily scale. It all…