If you want to run your model in the cloud, the Google Cloud ML Engine offers the power and flexibility of TensorFlow, scikit-learn, and XGBoost in the cloud. If this is not suitable, you can pick the API services that best fit your scenario. In Google Cloud ML, multiple APIs are available. These are classified into four major categories, as follows:

• Vision: The Cloud Vision API helps with image recognition and classification; the Cloud Video Intelligence API helps with scene-level video annotation; and AutoML Vision helps with custom image classification models.
• Conversation: Dialogflow Enterprise Edition builds conversational interfaces; the Cloud Text-to-Speech API converts text to speech; and the Cloud Speech-to-Text API converts speech to text.
• Language: The Cloud Translation API is used in language detection and translation; the Cloud Natural Language API is used in text parsing and analysis; AutoML Translation is used in custom domain-specific translation; and AutoML Natural Language helps in building custom text classification models.
• Knowledge: The Cloud Inference API derives insights from time series datasets.

There are also other services that are popular with developers, including Dialogflow and Wit.ai.