If your only motivation for using Google Go is webscraping, and you want to do you ML in python, I would recommend the following stack:

[Python requests for scraping data][1]

[MongoDB for caching data][2]

[pymongo for interfacing python and mongodb][3]

[scikit-learn for doing your machine learning][4]

This all happens in python and you can extend it multiple processors with 
[multiprocessing][5] or to multiple nodes with django


  [1]: http://docs.python-requests.org/en/latest/
  [2]: http://www.mongodb.org/
  [3]: http://api.mongodb.org/python/current/
  [4]: http://scikit-learn.org/stable/
  [5]: https://docs.python.org/2/library/multiprocessing.html