Apply AI: Update Your Resume Final Exam Answers

Apply AI: Update Your Resume Final Exam Answers

This final exam answer section provides complete, accurate, and well-structured responses to all questions related to using Artificial Intelligence (AI) to enhance and update a resume. It explains how AI tools assist in analyzing job descriptions, selecting powerful keywords, rewriting bullet points, improving grammar, and optimizing formatting for Applicant Tracking Systems (ATS). The content is designed to help learners understand how AI-driven techniques can transform an ordinary resume into a professional, targeted, and competitive document. These final exam answers ensure that students grasp the full value of integrating AI into the resume-building process and are prepared to apply these skills effectively in real job applications.

1. What are the main criteria you should consider when deciding between using a local (private) language model and a public chatbot for a task? Choose the 2 that are correct.

  • Privacy and data sensitivity of your information
  • Performance and capability of the model
  • The user interface design and color scheme of the tool
  • Whether the model was released in the past month

2. When selecting a local large language model to run on your computer, what are some hardware considerations, and why do they matter? Choose 3 that are correct.

  • The amount of RAM (memory) in your computer, because larger models require more memory to run smoothly without crashing other applications.
  • Whether your computer has a dedicated GPU or an Apple M-series chip, since these can significantly speed up how quickly the model generates responses.
  • The amount of storage space available on your hard drive, because the model’s processing speed depends primarily on storage capacity.
  • The capacity of your CPU, because more powerful processors will help the model run faster and handle more complex tasks.

3. If your local LLM is struggling with a complex task but you need to protect private data, what should you do? Choose 2 that are correct.

  • Break the complex task into subtasks and use a public chatbot only for the parts that do not require access to sensitive information.
  • Use your local LLM to identify and remove or mask sensitive information, then send only the non-sensitive parts to the public chatbot for processing.
  • Upload the entire document with all private information to the public chatbot, since it can handle more complex tasks.
  • Stop using language models entirely and do all the work manually, regardless of the task’s complexity or privacy requirements.

4. If you are unsure of the best prompts or task breakdown for your resume project, which processing approach can help you learn and improve your process more effectively?

  • Process a single section of your resume from start to finish before applying the process to other sections.
  • Use bulk processing to update all sections of your resume and related documents at once.
  • Skip the extraction and synthesis steps and just edit the entire resume manually.
  • Process all sections for each task before moving on to the next task, without reviewing or adjusting your process in between.

5. Why does it help to identify if certain steps in your workflow depend on earlier steps?

  • So you can save useful details or information during earlier steps that will be valuable for later tasks, such as customizing your resume or updating the skills section.
  • So you can skip any steps that seem time-consuming.
  • Because it allows you to delegate all steps to an automated tool without oversight.
  • Because it means you only need to focus on the final step and ignore the rest.

6. What are 2 ways that you can handle your source documents (notes, progress reports) safely with an LLM? Choose 2 that are correct.

  • Use a local LLM to review documents for sensitive information before sharing them with a public chatbot.
  • Upload all source documents directly to any public chatbot without reviewing for sensitive content.
  • Remove or mask confidential information from your documents before processing them with a public chatbot.
  • Trust that public chatbots will automatically protect your private information, so no extra steps are necessary.

7. What are some ways that you can evaluate the output of a language model? Choose 4 that are correct.

  • Ask the language model to perform a self-check or review of its previous output.
  • Accept all outputs as accurate without any further review.
  • Compare the model’s output to the original source documents to ensure accuracy.
  • Use diff tools to compare different versions line by line.
  • Organize the output in tables or spreadsheets to assist with systematic human review.

8. Why is section-by-section processing helpful compared to bulk processing of your resume? Choose 2 that are correct.

  • It helps prevent information from different sections or projects from getting mixed up, keeping details more accurate and relevant.
  • It allows you to focus on one section at a time, making it easier to review and improve the process before applying it to other sections.
  • It guarantees that all sections will be completed instantly with no manual effort.
  • It ensures your resume is automatically formatted for every employer’s requirements.

9. What options do you have when you want to redo a prompt? Choose 3 that are correct.

  • Start a new chat and re-enter the prompt from the beginning.
  • Edit the existing prompt and regenerate the response (for applications that have these features), replacing the previous output.
  • Branch or duplicate the conversation at the prompt you want to redo, creating a new version while keeping the original.
  • Wait for the model to automatically redo the prompt on its own without any user action.

10. What can you do if you want a language model to perform a task, but don’t know enough about the task to break it down into more detailed instructions? Choose 2 that are correct.

  • Ask the language model to explain the available choices for the task before proceeding.
  • Use meta prompting by asking the language model to help you write an effective prompt for the task.
  • Guess at the instructions and hope the model figures out what you want.
  • Avoid using the language model altogether if you aren’t already an expert in the task.

11. What are ways that you can take a hybrid approach to using a local LLM and public chatbot to perform a task? Choose 2 that are correct.

  • Anonymize sensitive information using the local LLM, then have the public chatbot perform the main task, and finally use the local LLM to de-anonymize the results.
  • Use the public chatbot to create a template (such as for formatting in HTML), then have the local LLM fill in your actual content using that template.
  • Always use only one tool for all tasks to avoid any workflow complications.
  • Share all your sensitive information with the public chatbot first, then let the local LLM review the output for privacy concerns.

12. When you are designing a workflow to perform a multi-step task, what are some things you can consider to guide your workflow design? Choose the 4 that are correct.

  • How to protect privacy, such as by anonymizing sensitive information before sharing with external tools.
  • How to optimize for speed, quality, and organization at each step of the process.
  • Prototype your workflow, critique the initial version, and iterate to improve it.
  • The colors and font styles used in the workflow diagram, since these directly impact performance.
  • Identify if later tasks depend on the outputs of earlier steps.

13. How can you make it easier for you to review a language model’s revision of your document (such as a resume)? Choose the 3 that are correct

  • Ask the language model to organize the original content, job requirements, its explanations, and suggested revisions in a table with row IDs.
  • Use a spreadsheet to store and review the table of revisions, making it easier to approve, reject, or modify each change individually.
  • Only look at the final revised document without comparing it to the original or reviewing intermediate steps.
  • Use a diff tool to compare line by line changes before and after the revision.

14. What are some things to be aware of when giving tasks of varying level of difficulty for language models to perform? Choose the 2 that are correct.

  • An “extract direct quotes” is an easier task for an LLM and can lead to fewer hallucinations compared to a “rewrite” task.
  • Putting multiple steps into a single prompt is a harder task for an LLM than breaking up tasks into separate prompts.
  • An “extract direct quotes” task is a harder task for an LLM and can lead to more hallucinations than a “rewrite” task.
  • Putting multiple steps into a single prompt is an easier task for an LLM than breaking it up into separate prompts.

15. What are some examples where your workflow is designed to optimize for speed, while still maintaining an adequate level of quality? Choose 2 that are correct.

  • Have the language model generate a tabular output with row IDs instead of manually adding row IDs in a spreadsheet.
  • Put a multi-step task into a single prompt instead of chaining multiple prompts in series.
  • Uploading monthly progress reports and performance review documents into a chatbot because chatbots run faster than a local language model.
  • Have the language model make final revisions of your resume to save time on the human review.