LLM Prompting Class | Dataford Interview Questions

Implement a class that initializes an LLM in the constructor, builds a prompt inside a method, and sends it to the LLM to produce structured output.

Asked in the Technical assessment (OA) stage. No framework; imports are prebuilt, with LeetCode-style function bodies.

For grading, implement structured_answer(request, llm_response). The function must build a prompt from request, parse the supplied LLM response as JSON, and return an object containing answer, confidence, and sources. Raise ValueError when the response is invalid JSON or misses a required field.

Input and Output

request is an object containing question and an optional context. llm_response is a JSON string. Return the parsed object when valid.

Constraints

request contains a string field named question.
request may contain an optional string field named context.
llm_response is a string.
A valid response must be a JSON object containing answer, confidence, and sources.

Function Signature

def structured_answer(request, llm_response):

Your question is LLM Prompting Class. Start with the requirements in the Question tab.

Run and submit as often as you like. When you're ready, talk me through your approach or go straight to the code.

Problem