How can i crawl the web in C# automatically without defining the fields in C#? Maybe with the help of Machine Learning?

I have e.g. 100 websites, which offer different products. I wanna have there content e.g label and price from the products but i don't wanna configure each web crawler manually with XPath or CSS Selector. I wanna crawl the whole page und use a Machine Learning package that analyze it. After that it will show me the label and product and detail fields of the products that i will write into database. Do you a technology i can use in C#? Is it possible and will be good enough to work?

Thanks !!