Week 14: Creating Professional Documents with AI

Learning objectives

By the end of this week, you will be able to:

1. Explain what makes a piece of writing sound and read as "professional," in terms of both tone and structure.
2. Describe the differences among a report, a proposal, and a set of meeting minutes, and identify a real situation where each one is the right document to write.
3. Draft a resume section with AI help while keeping every fact accurate, applying this course's verification habits.
4. Identify at least three common markers of generic "AI voice" in a piece of writing.
5. Apply a practical technique for rewriting an AI-voice paragraph into your own natural phrasing.
6. Give an AI tool a sample of your own past writing as a style reference, building on the "give examples" prompting technique from earlier in this course.
7. Use your screen reader's heading navigation, paragraph navigation, and Say All, or continuous reading, commands to review a long AI-drafted document efficiently before sending it.
8. Produce one professional document draft, critically edited for accuracy and voice, as this week's portfolio piece.

Lesson

Every AI tool you have used so far in this course, such as Claude, ChatGPT, and Gemini, is very good at producing a polished-looking paragraph quickly. That speed is genuinely useful, and this week leans into it directly. But speed creates a specific risk for professional writing: it is easy to accept a fast, confident-sounding draft without checking whether it is accurate, and without checking whether it actually sounds like you. This week teaches you to draft four common professional document types with AI help, a report, a proposal, meeting minutes, and a resume, and then teaches you, just as carefully, how to edit that AI draft so the facts are correct and the voice is your own.

What makes writing "professional"

A professional document is any piece of writing meant to be read at work: by a manager, a colleague, a client, a committee, or a future employer, rather than by a close friend. Professional writing is judged on two things: tone and structure. Tone means the overall feeling the words create, whether the writing sounds respectful, clear, and appropriate for the reader, or whether it sounds too casual, too stiff, or unclear. Structure means how the document is organized: whether the most important point comes early, whether related ideas are grouped together, and whether a reader can find what they need without hunting for it.

Professional tone sits in the middle of a range. On one end is the very casual style you might use texting a friend, full of slang, sentence fragments, and no explanation of background. On the other end is an overly stiff, old-fashioned style, full of long words chosen to sound impressive rather than to communicate clearly. Neither end serves a reader well. Professional writing lands in between: it is clear, direct, and free of unnecessary slang, but it does not use complicated words just to sound formal. A useful test is to ask yourself whether a busy colleague could read your document once, quickly, and understand exactly what you mean and what, if anything, you need from them.

Professional structure is just as important as professional tone, and it is the part writers most often skip when they are in a hurry. A well-structured professional document usually states its main point or purpose near the top, rather than making the reader wait until the last paragraph to find out why the document was written at all. It breaks information into short paragraphs, each covering one idea, rather than one long, unbroken block of text. And when the document is more than a page or so, it often uses headings to mark each major section, exactly the kind of heading structure this course has used in every week's lesson, so a reader using a screen reader, exactly like a sighted reader skimming visually, can move directly to the section they need using heading navigation rather than reading the whole document from the top every time.

AI tools are genuinely useful for both halves of this: they can suggest a clearer tone for a sentence that sounds too casual or too stiff, and they can suggest a more logical order for a document's sections. But an AI tool cannot know, on its own, who your specific reader is, what that reader already knows, or what outcome you actually want from the document. You still have to tell it that context, using clear-instructions and context-giving techniques from earlier prompting lessons, and you still have to judge, afterward, whether the tone and structure the AI produced actually fit your specific situation.

Four document types, and when to use each

This week focuses on four common professional document types. They share the tone-and-structure ideas from the last section, but each one has its own purpose, and knowing which one a situation calls for is itself a professional skill.

A report is a document that presents information or findings on a topic, usually to help someone else understand a situation or make a decision. A report is fact-focused: it describes what happened, what was found, or what the current status is, often ending with a short summary or a recommendation. A weekly project-status update, a summary of research on a topic, or a write-up of what a team accomplished this quarter are all reports. You reach for a report when your goal is to inform: someone needs to understand a situation clearly, and your document's job is to give them that understanding.

A proposal is a document that argues for a specific course of action, usually asking someone to approve it, fund it, or agree to it. Unlike a report, a proposal is persuasive rather than purely informative: it names a problem or an opportunity, describes a specific solution, and explains the benefits of that solution clearly enough that the reader is convinced to say yes. A pitch to a manager for a new tool or process, a request for funding for a project, or a written case for changing how a task is done are all proposals. You reach for a proposal when your goal is to convince: you have a specific ask, and your document's job is to make the case for it.

Meeting minutes are a written record of what happened during a meeting: the decisions that were made, the action items that were assigned, and who is responsible for each one. Minutes are deliberately plain and factual, not persuasive and not full of opinion. A good set of minutes lets someone who missed the meeting understand exactly what was decided and what happens next, without needing to have been in the room. You reach for meeting minutes immediately during or right after a meeting, specifically to create that shared, accurate record for everyone involved, including people who were not there.

A resume is a document that summarizes a person's work history, education, and skills, written specifically to help an employer decide whether to interview that person for a job. Other material in this course covers resumes from a career-strategy angle: which skills to highlight and how to research a target job. This week's focus is narrower and more practical: how to actually write resume content with AI help, section by section, while keeping every fact completely accurate. You reach for resume writing whenever you are applying for a new role, or whenever you want an up-to-date document ready before an opportunity appears.

Notice what these four document types have in common, and where they differ. All four benefit from the clear tone and organized structure from the previous section. But a report informs, a proposal persuades, minutes record, and a resume summarizes and presents a person favorably. Naming which one of these four jobs your document needs to do, before you write a single word, is the single most useful planning step you can take, and it is also the single most useful piece of context to give an AI tool when you ask it to help you draft.

Drafting a resume section by section with AI help

A resume is easiest to draft, with or without AI help, one section at a time rather than all at once. The common sections are contact information, a short summary or objective statement near the top, work experience, skills, and education. Working section by section keeps each piece of writing short and specific, which makes both drafting and checking far easier than trying to write, or trying to verify, an entire resume in one pass.

An AI tool can genuinely help with resume writing in several ways. Given a plain description of what you did in a job, it can suggest stronger, more specific action verbs. It can tighten a wordy sentence into a short, punchy bullet point, the format most resumes use for work experience. And it can help you phrase the same underlying accomplishment slightly differently for different job applications, a skill that later, more career-focused material in this course builds on directly when it covers tailoring a resume per job.

Here is exactly where this course's earlier hallucination and verification lesson becomes essential. Hallucination, as this course defined it earlier, is when an AI tool states something confidently that is not actually true. On a resume, hallucination often shows up as quiet embellishment: you describe a task in your own plain words, and the AI's polished version accidentally makes it sound bigger, more senior, or more precise than what actually happened. A resume with even one inaccurate claim is a serious problem, not a minor one, since an employer may ask about it directly in an interview, and because honesty was named as a core ethical habit for AI use earlier in this course.

The fix is a simple, repeatable habit: read every single AI-suggested resume sentence against your own memory of what you actually did, word by word, and change anything that overstates it, before it goes anywhere near a real application. Consider this example. Your own plain description of a task might be: "I helped organize the team's files and answered phone calls." A first AI draft, asked to make that sound stronger, might come back as: "Managed the department's entire records system and served as primary point of contact for all client communications." Read carefully, that AI version claims two things that may not be true: that you managed the entire system, rather than helping organize it, and that you were the primary contact for all communications, rather than one person who answered some calls. A properly verified, accurate bullet keeps the stronger language the AI style found, while keeping the facts exactly as they happened: "Helped organize and maintain the department's shared files, and regularly answered incoming client phone calls." That version is still clear, active, and professional. It is simply true, which matters far more than sounding impressive.

Finding and removing generic "AI voice"

Beyond accuracy, there is a second, separate problem worth learning to catch: AI-drafted writing often has a recognizable style of its own, sometimes called "AI voice." AI voice is not inaccurate, exactly. It is generic: it reads as if it could have been written by anyone, about almost anything, rather than sounding like a specific person with a specific point to make. A document written entirely in AI voice can pass a fact-check and still fail a reader, because it does not sound like you, and a document under your name should sound like you wrote it.

Three markers show up again and again in AI voice, and learning to notice all three is the whole skill. The first is vague enthusiasm: phrases like "I am thrilled to share" or "This is an exciting opportunity" that express a feeling without giving any actual information. The second is repetitive sentence structure: several sentences in a row starting the exact same way, such as beginning paragraph after paragraph with "In today's fast-paced world" or "It is important to note that." The third is overused transition words: words like "moreover," "furthermore," and "additionally" appearing so often that they stop guiding the reader and start feeling like padding.

Here is a short example that shows all three markers at once, followed by an edited version with the same information in the writer's own words. The AI-voice version: "I am thrilled to share the results of this quarter's project. In today's fast-paced world, staying organized is more important than ever. Furthermore, our team successfully completed the report ahead of schedule. Moreover, this achievement demonstrates our strong commitment to excellence." Notice the vague enthusiasm in the first sentence, the generic opening cliche in the second sentence, and the two overused transition words stacked back to back. Here is the same information, edited to remove all three markers: "This quarter's project finished two weeks ahead of schedule. The team split the report into sections early, which kept everyone on track even when the deadline moved up. That head start is the main reason we came in early." The edited version is shorter, more specific, and sounds like a person describing what actually happened, rather than a generic announcement that could apply to almost any project.

A practical technique makes this editing manageable rather than overwhelming. First, read the AI draft aloud, or listen to it with Say All, and mark any sentence that expresses a feeling without giving a fact, such as "exciting" or "thrilled," without saying exactly what happened. Second, scan for sentences that all start the same way, and rewrite at least every other one so they begin differently. Third, search for the most commonly overused transition words, "moreover," "furthermore," "additionally," and "in today's fast-paced world," and either delete them outright or replace them with a specific detail instead. This three-part check takes only a few minutes once it becomes a habit, and it is the single most effective tool this week teaches for making an AI-drafted document sound like your own writing rather than like a template.

Giving AI a style reference from your own writing

The editing technique in the previous section fixes AI voice after the fact. An equally useful technique prevents much of it from appearing in the first place: giving the AI tool a sample of your own past writing as a style reference before it drafts anything new.

An earlier lesson in this course taught the "give examples" prompting technique: showing an AI tool a concrete example of the kind of output you want, rather than only describing it in the abstract, because a good example often communicates a style faster and more precisely than a paragraph of instructions. Applying that same technique to your own voice is simple in practice. Find a short piece of writing you produced yourself in the past, an old email, a short report, or a paragraph you are proud of, and paste it into the conversation before asking the AI tool to draft your new document. Then ask directly: "Here is a sample of my own writing. Please match this tone and sentence style when you draft the new document below," followed by your actual request.

This works because an AI tool responds strongly to the examples it is given, which means a style sample tends to steer its word choice, sentence length, and level of formality much more precisely than a general instruction like "sound professional but casual" ever could on its own. It also builds directly on the system-prompt and custom-instructions idea from earlier lessons. Just as a saved system prompt or a custom instruction tells an AI tool your standing preferences so you do not have to retype them every conversation, you can save a short style reference of your own writing as part of a reusable instruction document, the same kind of file an earlier lesson taught you to build, so every future document-drafting conversation starts already knowing roughly how you write.

One caution matters here, and it echoes the honesty and privacy points made earlier in this course: only paste writing you are comfortable sharing with that AI tool, and never paste a document that contains someone else's private or confidential information just to use as a style sample. A short paragraph you personally wrote, with any sensitive names or details removed, is almost always sufficient to capture your voice, and it avoids that risk entirely.

Reading a long AI-drafted document efficiently

Editing an AI-drafted document well requires reading the whole thing carefully, at least once, before it goes anywhere. For a short paragraph, reading start to finish with Say All, or your screen reader's continuous reading command, is enough. For a longer report or proposal, several pages long, reading strictly start to finish can be slow, and it makes it easy to lose track of where you are in the document's overall structure. Three review techniques, used together in a deliberate order, make reviewing a long document far more efficient.

Start with heading navigation, using your screen reader's heading navigation command to jump from heading to heading, the same skill this course has used since early on to move around a webpage. Ask your AI tool to draft the document with clear headings for each major section, exactly as this course's own lessons are structured, and then use heading navigation to jump through those headings first, before reading anything in full. This first pass gives you the document's overall shape: how many sections it has, and whether that overall structure actually matches what you asked for, before you invest time reading every sentence.

Next, use paragraph navigation within a section that needs closer attention. In a browser or a text editor, Ctrl plus Down Arrow moves forward one paragraph at a time, and Ctrl plus Up Arrow moves back one paragraph, letting you step through a section's ideas one at a time without committing to reading the entire document straight through. This is especially useful when you already know, from the heading pass, which one or two sections need the closest fact-checking, such as the work-experience section of a resume from earlier in this lesson.

Finally, once you have edited a section, read the finished result in full with Say All, or your screen reader's continuous reading command, the same technique this course has used throughout for a full, uninterrupted read-through. This full read-through is the right way to catch how a document actually sounds when read start to finish, including the AI-voice markers from earlier in this lesson, which are often easier to notice by ear, in the natural flow of listening, than by scanning text visually or jumping between individual sentences. Screen readers differ in exactly which keys perform heading navigation and continuous reading, so check your own screen reader's documentation or shortcut list if you are not already sure of them.

Used in this order, heading pass first, then paragraph-by-paragraph editing of the sections that need it, then a full continuous read at the end, this three-step method lets you review a long AI-drafted document thoroughly without needing to read every single sentence three separate times. It is also exactly the kind of exploration habit this course has taught for any unfamiliar interface: use the structure the document or the tool already gives you, headings and paragraphs, rather than fighting through it in one long, undifferentiated block.

Where this fits in the bigger picture

This week's four skills, recognizing what makes writing professional, choosing the right document type, verifying every fact on a resume, and catching AI voice, are really one skill practiced four different ways: reading AI output critically instead of accepting it automatically. That is the same underlying habit this course built earlier for AI output in general, now aimed specifically at the professional documents that make up so much of paid work. It is also a direct, hands-on application of the AI-assisted writing skill area named earlier in this course as one of the skills employers value most, and it uses the prompting techniques, giving examples and setting standing instructions, that earlier lessons in this course built.

Later material in this course builds on the very document you draft this week, turning your edited draft into a properly formatted, accessible file, and teaching you to check that file for accessibility before you send it to anyone else. This week's job is getting the words right. Getting the finished file right comes later.

Key terms from this week

Use this list to review the vocabulary introduced in this lesson before you start the exercises and the test.

Report

A document that presents information or findings on a topic, usually to inform a reader or help them make a decision, often ending with a summary or recommendation.

Proposal

A document that argues for a specific course of action, naming a problem, a solution, and its benefits, in order to persuade the reader to approve it.

Meeting minutes

A plain, factual written record of a meeting's decisions and action items, created so someone who was not present can understand exactly what happened.

AI voice

A generic writing style common in AI-drafted text, marked by vague enthusiasm, repetitive sentence structure, and overused transition words, that reads as if it could have been written by anyone.

Style reference

A sample of a person's own past writing given to an AI tool so it can match that person's tone and sentence style when drafting something new.

Keyboard-only exercises

These exercises use only your keyboard. Screen reader commands are described in general terms below; substitute your own screen reader's equivalent command where needed.

Exercise 1: Draft one professional document with AI help

1. Open an AI chatbot such as Claude, ChatGPT, or Gemini.
2. Choose one document type from this week's lesson: a short report, a short proposal, a set of meeting minutes, or one section of a resume.
3. Before asking for a draft, type a short piece of your own past writing into the same conversation as a style reference, following this week's lesson on giving AI a style sample.
4. Ask the AI tool to draft the document, naming the document type clearly and matching the tone of your style sample, using clear-instructions and context-giving techniques from earlier in this course.
5. Read the draft in full using Say All, or your screen reader's continuous reading command.
6. Save the draft as a plain text or Markdown file, using file-saving skills from earlier in this course, with a clear file name such as document-draft-v1.

Exercise 2: Review the draft's structure and hunt for AI voice

1. Open the draft from Exercise 1 in your document editor or browser.
2. Use your screen reader's heading navigation to jump through the document's headings first, checking that the overall structure matches what you expected.
3. Use Ctrl plus Down Arrow to move paragraph by paragraph through one section, and Ctrl plus Up Arrow to move back, listening for the three AI-voice markers from this week's lesson: vague enthusiasm, repetitive sentence starts, and overused transition words.
4. Mark or note, in a new paragraph at the bottom of the file, at least two sentences that show one of these three markers.
5. Save the file, keeping your notes together with the original draft.

Exercise 3: Edit for accuracy and voice, then compare before and after

1. Open the draft and notes from Exercise 2.
2. Applying this course's verification habits, check every specific fact, name, date, or claim in the draft against what you actually know to be true, and correct anything that is exaggerated or inaccurate.
3. Rewrite each sentence you marked in Exercise 2 in your own natural phrasing, removing the AI-voice marker you identified, following this week's three-part editing technique.
4. Read the fully edited version in full with Say All, or your screen reader's continuous reading command, and confirm it sounds like your own writing.
5. Save the edited version as a new file, such as document-draft-v2, keeping the original draft from Exercise 1 as your "before" version for comparison.

Portfolio project: My AI-Drafted Professional Document

This week's portfolio piece is a complete before-and-after example of AI-assisted professional writing, along with a short reflection explaining your editing choices. Keep this piece; later course work will build directly on it.

1. Choose one document type: a report, a proposal, a set of meeting minutes, or one section of a resume.
2. Draft the document with an AI tool, using a short sample of your own past writing as a style reference, following Exercise 1.
3. Critically edit the draft: verify every fact against what actually happened, following this course's verification habits, and remove generic AI voice using this week's three-part technique.
4. Keep both versions: the original AI draft, labeled "Before," and your critically edited version, labeled "After."
5. Write a short reflection, three to five sentences, describing what you changed and why, naming at least one factual correction and at least one AI-voice edit you made.
6. Save all three parts, "Before," "After," and your reflection, together in one file or one folder named my-ai-drafted-professional-document, in the same portfolio folder you have used throughout this course.

Completion checklist:

• You chose one document type and named it clearly.
• You gave the AI tool a style reference from your own past writing before drafting.
• You checked every fact in the draft and corrected any inaccuracy.
• You identified and rewrote at least two AI-voice sentences.
• You kept the "Before" and "After" versions together, along with your written reflection.
• You saved the finished portfolio piece in your portfolio folder.

Weekly test

Answer key

Each answer below includes a one-sentence explanation drawn from this week's lesson.