AutoML in Auto-sklearn

My research is focusing on improving AutoML, especially on the pre-processing part. It is stated in some articles that at the moment, data pre-processing is not handled very well by any of the AutoML tools and still requires considerable human intervention. I propose to use multimodal data fusion by using deep learning to convert every data type to numerical which in pre-processing part. The AutoML that I want to focus on is Auto-sklearn which uses python coding. I am a beginner in python by the way.

So, I need to have systematic coding source in python of auto-sklearn and also coding source in multimodal deep learning to understand how to integrate both frameworks. If somebody can be interested to approach me, please let me know.