Abstract: A system and methods are described for retrieving, displaying, and/or analyzing law firm analytic data in response to natural language queries. The system, at an interface or bot, receives queries as plaintext messages requesting information about or maintained by a law firm. The system generates intent scores for the message for each of multiple intents and determines an intent for the message based on a highest generated score exceeding a threshold. The system extracts entities from the message, applies an algorithm to the determined intent and extracted entities, and identifies a data source from which to retrieve information. The system then retrieves and displays information from the identified data source.