Function calling
Function calling lets an AI model trigger external tools or actions, like looking up an order, by returning a structured request the system can run.
Function calling is a capability that lets an AI model do more than generate text. Instead of only replying in words, the model can return a structured request that tells the surrounding system to run a specific tool or action, such as querying a database, checking an order status, or creating a record.
The model does not run the action itself. It produces a structured output, often as JSON, that names the function to call and the values to pass. The application runs the function and feeds the result back to the model, which then uses it to form a final answer. This lets an AI assistant work with live data and real systems.
In context
In customer support, function calling is what lets an AI assistant move from answering general questions to handling real tasks. When a customer asks where their order is, the assistant can call an order lookup function, receive the live status, and report it back in plain language rather than guessing.
This same pattern supports actions like checking account details, processing a return, updating a shipping address, or creating a ticket. Function calling connects the assistant to helpdesk and back office systems so it can resolve requests end to end instead of only explaining what a customer should do.
How Auralis uses Function calling
Auralis uses function calling so Autopilot and Assist can connect to your helpdesk and back office systems, look up live order and account data, and resolve customer requests rather than just describe the steps.
