Should I use Zorroin my Python solution ?!

Hello,

I recently discovered Zorro Trading which seems to be an amazing trading tool.
I am wondering how I could use it and integrate it in my Python algo trading solution.
The ting is that my solution is 90% done so I'm not sure it worth investing time in plugging in new components.

Below is a small description of my trading solution.
Could you recommend the steps at which it is the most interesting to use Zorro in replacement/addition to my solution?

- end to end solution developed in Python
- fetch raw data from Yahoo Finance API, soon to be replaced with direct broker's data (Oanda and Interactive brokers)
- computation of "technical indicators" some of them being homemade and complex like trendlines, support/resistance
- home made Machine learning engine (ensemble learning)
- home made optimization engine that basically brute force millions of parameters related to indicators and ML engine to optimize results
- rudimentary home made back testing (simply takes the forecasts + trading rules and checks the expected returns on the out of sample dataset)
- to be done: portfolio balancing, order execution with the API of brokers

Any suggestion is welcome!

If your solution is 90% ready, I would finish it with Python, and use Zorro only for optimizing, backtesting, and live trading it.

JCL answering me, I'm honored!
I was reading your Financial Hacker blog earlier, loved it!

I thought about something.

My Python code already produces forecasts and for each day a decision to buy or sell.
Could I pipe these buy/sell decisions (and a few trading rules) into Zorro to deliver a backtest ?

Most back testers are rules driven (you code your strategy in them).
But I don't have a strategy, I have an ML black box giving me forecasts...

Sure, let Zorro call your Python function and use the returned predition for a trade signal. You need Zorro 2.50 when your Python code is 64 bit.

http://manual.zorro-project.com/python.htm

Thanks, I now have a good reason to try Zorro!