We are all paying attention to Machine and Deep Learning. Due to the impressive learning capabilities of these models, we are now building things that we thought were not possible before, from smarter text translation to self-driving cars.

Now, the question is: can we use AI, machine and deep learning to help software developers in building software? Can models learn how to find bugs? Can models learn what code quality means to you and your team and recommend refactoring? Yes, they can!

In this talk, I will walk you through my adventures in applying machine learning to different SE problems, and the challenges I observe whenever we deploy them in software companies.

Video (in english)

Video (in pt-br)


  • ThoughtWorks
  • Zühlke
  • University of California, Davis
  • Universidade Federal do Ceará, campus Quixadá
  • Lambda3
  • RunaHR


  • https://www.mauricioaniche.com/publications/log-placement-recommendation/
  • https://www.mauricioaniche.com/publications/learning-off-by-one-mistakes/
  • https://www.mauricioaniche.com/publications/search-based-remodularization/
  • https://www.mauricioaniche.com/publications/ml-and-refactoring/
  • https://www.mauricioaniche.com/publications/anomaly-detection-cps/
  • https://www.mauricioaniche.com/publications/evosql/
  • https://www.mauricioaniche.com/publications/passive-learning-applied/