AI-integrated logistic systems and methods

A method to control artificial intelligence (AI) human-machine interaction within a transportation management system, the method comprising:

• generating, with a processing system comprising one or more electronic processors, a user interface to receive a user query related to vehicle transportation;
• receiving, with the processing system, via the user interface, the user query related to vehicle transportation;
• providing, with the processing system, the user query to an artificial intelligence (AI) system including one or more machine learning models, the AI system to pre-process the user query to transform the user query to a processed user query, wherein the processed user query is the user query augmented with a configuration file that is indicative of an intent of the user query and an entity of the user query;
• receiving, with the processing system, from the AI system, the processed user query;
• providing, with the processing system, the processed user query to the AI system, wherein the AI system is to access, based on the processed user query, transportation data from a database that stores information related to vehicle transportation, wherein the transportation data is related to the intent of the user query as indicated by the configuration file;
• receiving, with the processing system, from the AI system, the transportation data;
• providing, with the processing system, the transportation data and the user query to the AI system, wherein the AI system is to determine an automated answer to the user query based on the transportation data;
• receiving, with the processing system, from the AI system, the automated answer to the user query;
• providing, with the processing system, the user query and the automated answer to a large language model (LLM) of the AI system to determine a validity of the automated answer with respect to the user query;
• responsive to the LLM determining that the automated answer is an invalid answer to the user query:
• providing, with the processing system, the user query to the AI system to determine a recommended user query based on an intent of the user query, wherein the recommended user query is to remediate a cause of the automated answer being an invalid answer to the user query;
• and
• transmitting, with the processing system, the recommended user query to a user device for display via the user interface;
• receiving, with the processing system, a second automated answer for the recommended user query;
• transforming, with the processing system, the second automated answer for the recommended user query into a human readable format as a response to the recommended user query;
• updating, with the processing system, the user interface to include the response to the recommended user query as an updated user interface;
• and
• transmitting, with the processing system, the updated user interface to the user device for display using the user interface.

A non-transitory computer-readable medium storing instructions that, when executed by one or more electronic processors of a processing system, cause the processing system to perform operations comprising:

• generating a user interface to receive a user query related to vehicle transportation;
• receiving, via the user interface, the user query related to vehicle transportation;
• providing the user query to an artificial intelligence (AI) system including one or more machine learning models, the AI system to generate an automated response to the user query based on transportation data that is accessible from a database and is relevant to answering the user query;
• providing, with the processing system, the user query and the automated response to a large language model (LLM) of the AI system to determine a validity of the automated response with respect to the user query;
• responsive to the LLM determining that the automated response is an invalid response to the user query:
• providing the user query to the AI system to determine a recommended user query based on an intent of the user query, wherein the recommended user query is to remediate a cause of the automated response being an invalid response to the user query;
• and
• transmitting the recommended user query to a user device for display via the user interface.