resolveCustomer
A Python utility function defined in the article tutorial that wraps the M|BOX fuzzy matching logic. It is intended for AI agents to invoke in order to perform deterministic customer identity resolution, returning candidate records and field-level scores.
Timeline (last 90 days)
The resolveCustomer function takes four parameters—first name, last name, company, city—internally calls index.match to return the top three candidates with their field-level scores, and empty-string fields are excluded from the total score calculation.
The tutorial wraps resolveCustomer as a function tool of the OpenAI Agent SDK; the SDK automatically generates the JSON schema from the docstring, and each parameter becomes a string property.
The Agent's system instructions require: always call resolveCustomer, never guess a customer ID, pause and ask for more detail when the top score is below 80 or two candidates are within 10 points of each other, and the model temperature is set to 0.
All Facts (3)
The resolveCustomer function takes four parameters—first name, last name, company, city—internally calls index.match to return the top three candidates with their field-level scores, and empty-string fields are excluded from the total score calculation.
50%UnverifiedThe tutorial wraps resolveCustomer as a function tool of the OpenAI Agent SDK; the SDK automatically generates the JSON schema from the docstring, and each parameter becomes a string property.
50%UnverifiedThe Agent's system instructions require: always call resolveCustomer, never guess a customer ID, pause and ask for more detail when the top score is below 80 or two candidates are within 10 points of each other, and the model temperature is set to 0.
50%