Module 3

Writing with Artificial Intelligence

Drafting professional business correspondence—such as executive updates, client emails, and formal project memos—often consumes hours of administrative time. Generative AI tools serve as conversational drafting partners that convert rough outlines, bullet points, and raw meeting notes into structured, professional prose.

In business operations, speed must never compromise clarity or organizational voice. Using AI for first drafts allows you to overcome writer's block, establish consistent formatting, and quickly adapt communications for different audiences, such as senior leadership or cross-functional team members.

Guided Walkthrough: Drafting a Project Status Memo

To produce an effective first draft, supply clear context, the target audience, the intended tone, and source facts in your prompt. Consider this workplace scenario where a project coordinator needs to draft an internal update:

You are an executive communications assistant. Draft a 3-paragraph project update memo to departmental stakeholders based on the following notes:
- Project: Customer Portal Refresh
- Milestone: Phase 1 testing completed on schedule (April 12)
- Challenge: Vendor API documentation was delayed by two days, but the team adjusted sprint tasks with zero launch slippage.
- Next Step: User acceptance testing begins Monday.
Tone: Confident, clear, and professional. Use concise business language.

The AI generates a coherent initial draft within seconds. You can then refine the text using iterative prompts. For instance, follow up with: "Condense paragraph 2 into two sentences and format the next steps as a bulleted checklist." This iterative conversation tailors the message to exact organizational standards.

Common Beginner Mistake: Passive Acceptance

A frequent misconception among new AI users is assuming that a polished, grammatically correct draft is factually accurate and ready to send. AI models predict words based on probability; they do not know whether your project milestone genuinely met its deadline or whether company policy permits certain phrasing. Never copy and paste AI-generated correspondence directly into an outgoing communication without performing a careful human review.

Practice Checkpoint

Before moving forward, practice identifying the core inputs of your correspondence: role, audience, key facts, and tone constraints. In the upcoming lab, you will apply this technique to draft and refine a complete operational memo using browser-based AI tools.

Understanding Data Cleaning and AI Verification

Spreadsheets are central to business operations, but raw operational data often arrives disorganized. Inconsistent date formats, erratic letter casing, trailing whitespace, and mismatched categories prevent accurate analysis. Data cleaning is the process of detecting and correcting corrupt, inaccurate, or improperly formatted records within a dataset.

Conversational AI platforms can analyze messy sample entries and recommend step-by-step cleanup techniques, transformations, or native spreadsheet features (such as Text to Columns, Flash Fill, or TRIM functions). However, working with business data requires an active human-in-the-loop workflow—an operational safeguard where human judgment validates every AI suggestion before changes are committed to production files.

Workplace Context and Privacy Safeguards

When using browser-based AI assistants to solve data quality issues, data governance comes first. Never paste real customer records, employee identifiers, or proprietary financial metrics into public AI tools. Instead, extract two or three rows of anonymized or synthetic mock data that illustrate the structural issue.

Consider this realistic scenario: An administrative coordinator receives a messy list of regional office locations where phone numbers and state codes follow inconsistent formats:

Raw Entry 1: " austin , TX - (512)5550192 "
Raw Entry 2: "dallas, texas -- 214-555-0144"
Raw Entry 3: "HOUSTON, Tx : 713.555.0188"

You can prompt the AI assistant:

I have spreadsheet rows with inconsistent formatting like the three sample rows below. 
Provide a clean 3-column table (City, State Postal Code, Standard Phone Format: (XXX) XXX-XXXX) showing how these should look. Then provide the standard spreadsheet functions or features I can use to clean the entire column myself.
[Sample rows inserted here]

The assistant demonstrates the transformed target output and provides standard spreadsheet logic without ever viewing sensitive business records.

Common Mistake: Blind Bulk Transformations

A critical mistake is pasting hundreds of uninspected rows into an AI tool and pasting the output straight back into a master worksheet. AI models may silently alter numbers, drop leading zeros from postal codes, or misinterpret ambiguous abbreviations (such as converting "CA" to "Canada" instead of "California"). Always review sample outputs against your original source records to confirm data integrity.

Preparing for Practice

Maintaining data hygiene protects business decisions from faulty reporting. In the next lesson, you will learn how to turn clean data structures into dynamic formulas and executive summaries.

Constructing Formulas and Summaries with AI

Modern knowledge workers frequently need complex spreadsheet formulas to calculate metrics, cross-reference tables, and aggregate results for leadership. While building nested functions from scratch can be challenging, generative AI excels at translating plain-language business requests into precise spreadsheet syntax.

A formula prompt describes your table structure, column headers, and desired mathematical or logical outcome. Instead of struggling to memorize syntax for functions like XLOOKUP, SUMIFS, or COUNTIF, you describe your goal, and the AI writes the formula along with an explanation of each argument.

Guided Walkthrough: Generating an Aggregation Formula

Suppose you maintain an event registration tracking sheet with the following setup:

• Column A: Attendee Name
• Column B: Department (for example, Marketing, Sales, Operations)
• Column C: Registration Fee (numerical dollar value)
• Column D: Attendance Status (Confirmed, Pending, Cancelled)

You need to calculate the total registration revenue generated exclusively by confirmed attendees from the Sales department. You can provide this structured prompt:

I am working in a standard spreadsheet application. 
Sheet structure:
- Column B: Department
- Column C: Registration Fee
- Column D: Attendance Status

Write a formula that sums the values in Column C only if the Department in Column B is "Sales" and the Attendance Status in Column D is "Confirmed". Explain how the formula arguments work so I can adapt cell ranges if needed.

The AI model returns the recommended solution:

=SUMIFS(C:C, B:B, "Sales", D:D, "Confirmed")

It also explains that SUMIFS places the sum range (C:C) first, followed by each criteria range and condition pair. This breakdown allows you to verify the logic and adjust cell references (such as changing column references to specific ranges like C2:C150) to fit your sheet.

Creating Structured Tabular Summaries

Beyond isolated formulas, AI can suggest layout structures for summary tables and executive dashboards. When you describe the questions stakeholders need answered—such as quarterly budget variance or department-level headcount—the model can outline the exact rows, columns, and metric formulas required to build a clean report.

Common Beginner Mistake: Range and Syntax Mismatches

A common beginner mistake is copying an AI-generated formula directly into a spreadsheet without verifying cell references or localized formula separators. If your data begins in row 2 due to headers, using whole-column references or mismatched ranges can trigger #VALUE! or circular reference errors. Always match the AI formula's inputs to your specific sheet geometry.

Lab Transition

You are now ready for hands-on practice. In the upcoming lab exercise, you will open your browser AI tool, craft formula prompts, and generate structured summaries using realistic business scenarios.