I like how easy it was to get up-and running with sigopt, there are no fancy features or confusing interfaces so you can learn quickly without needing an expert in machine learning/data science behind every feature! There's nothing i dislike about this software but if anything they could add more tutorials for novice users who don't have experience using python libraries such as pandas etc., This has helped us gain insights into our data set by enabling quick prototyping of models which allows me (a non ML PhD)to build faster than my colleagues at work - just because we're all working from home doesn;t mean everyone needs access too expensive tools!! We used ROC curves when building predictive algorithms around customer behaviour patterns within ecommerce websites where customers tend not spend much time browsing product pages before making purchase decisions compared to looking through reviews sections first. The model achieved AUC scores over 0.8 once trained correctly meaning good accuracy predicting future purchases vs previous purchasing habits across different products categories. Using TensorFlow API along with Python gives great flexibility allowing one SigOpt has helped us solve problems in a variety of AI problems. We use it to solve machine learning problems, but we have also used it to solve optimization problems. It's a great platform. I don't have any complaints. I think it's a great tool. We've had success solving a variety of problems with SigOpt. We've solved optimization problems, but we've also solved machine learning problems. We've solved a variety of problems.