We don't publish
your competitive advantage.

IFWhen using LlamaIndex's structured LLM, the response may fail to parse into a Pydantic model, causing AttributeError on model_dump_json().

THENImplement a retry mechanism that catches Pydantic validation errors and re-generates the structured output up to N times. This ensures the output conforms to the expected model before proceeding.

IFWhen using Structured LLM in llama_index, the LLM response may be a plain string instead of a Pydantic model, causing an AttributeError: 'str' object has no attribute 'model_dump_json'.

THENImplement a retry mechanism that catches Pydantic validation errors during structured output parsing. After receiving the LLM response, attempt to parse it into the Pydantic model. If parsing fails, re-prompt the LLM with an error message asking for valid output. Limit retries to a maximum number (e.g., 3) and raise a custom exception if all fail. This approach mirrors the ReflectionWorkflow pattern used in llama_index examples.

IFStructured LLM output fails to parse into Pydantic model, returning a raw string and raising AttributeError on model_dump_json() call.

THENImplement a retry mechanism that catches Pydantic validation errors and re-calls the LLM with the same prompt, optionally including the error message as feedback. This ensures the output conforms to the expected schema before proceeding, similar to the retry logic used in the ReflectionWorkflow.