I am building a website in which the client enters his job role that he is applying to, and then a few questions appear to them based on the role. Then they can answer the questions and get review for the answers one-by-one.

The first prompt is:

const completion = await openai.chat.completions.create({
      messages: [
        {
          role: "system",
          content:
            "Ask me 5 questions (without any uneccessary words) that will help me get ready for my job interview in Software Engineering. I will try to answer them one by one. Then you will tell me how well I answered it.",
        },
      ],
      model: "gpt-4",
    });

Then for each answer:

const reviewCompletion = await openai.chat.completions.create({
      messages: [
        {
          role: "system",
          content:
            "You are a helpful assistant that reviews answers to interview questions and provides constructive feedback.",
        },
        { role: "user", content: `Question: ${question}` },
        { role: "user", content: `Answer: ${answer}` },
      ],
      model: "gpt-4",
    });

Is there a better way to achieve what I want that will give me more accurate reviews?