The platform was developed to handle multiple agent datasets within a single user login i.e you can add training data for any number of agents. The main focus of Alter NLU is to overcome all the roadblocks involved in building a stable and good quality training data. To achieve the same, it is segregated into 3 parts: Intents - Intents help you recognize what your users want to say. Entity - Entities describe the piece of information you would want to extract from the expressions/messages of the user. It identifies ‘ things’ that your users mention. Reports on training dataset quality - The platform did not just help train the data, but also provided AI-based real-time ‘Reports’ where it listed out the issues and recommended improvements such as  an in-depth analysis and alerts for the intents and entities that need more training pointing out what exactly needs to be addressed. This helps in maximizing the accuracy of the agents’ response.Cost factor: It is not easy to build a NLP trained chat assistant for those who do not come from a technical background. Heavenly dependent on either having an in-house Machine Learning team consisting of engineers, data scientists, product team or they will have outsource the product entirely