Proving Your AI-Assisted VA Skills

Capstone Brief: Meet the Client

1. A Real Client Does Not Give You One Perfect Prompt

In real VA work, information may arrive through a client introduction, an email, a process document, a spreadsheet, a meeting transcript, customer messages, screenshots, and scattered follow-up messages. Your first responsibility is to understand the account. Do not immediately start asking AI to produce something — first understand: who is the client? What do they need? What information has been provided? What requirements or constraints matter?

2. Meet BrightPath Learning Co.

For the capstone, you'll work with the fictional client you've used throughout the course — BrightPath Learning Co. You already know Sarah and the BrightPath environment. This time, the materials arrive together as a client handoff.

3. Do Not Treat Every Piece of Information the Same Way

As you review the materials, identify: Client Requirements (what does Sarah actually want?), Source Information (what facts or materials are you expected to work from?), Constraints (communication preferences, existing processes, approved information, deadlines, privacy requirements), Open Questions (what remains unclear?), Sensitive Information (what needs additional care?), and Deliverables (what actual work needs to be produced?).

4. Your First Job Is Understanding

Before producing anything, create a clear picture of the assignment. You should be able to answer: what does Sarah need? What are the separate tasks? What information supports each task? What should I clarify? What should I avoid assuming? Where might AI help? Where should I work without AI? This is the same discipline you developed earlier — now applied across an entire client account.

5. PRACTICE: Sort the Client Materials LEARNING ARTIFACT

Using the BrightPath Client Handoff Pack, for each document identify whether it's primarily a Client Requirement, Source Information, a Constraint, an Open Question, Sensitive Information, or Background Context. If a document contains information that would require special handling (like the customer's contact details), identify that specific information rather than assuming the entire document is sensitive — most of this handoff is ordinary business content. Some information may serve more than one purpose — explain your reasoning when necessary.

6. PRACTICE: Identify the Work

From the client handoff, identify the major deliverables — don't start producing them yet.

7. PRACTICE: Where Should AI Help?

For each deliverable, identify AI Can Help / AI Can Help With Caution / Primarily Human, and explain why. Remember the principle from Lesson 1.1: classify the specific task, not the whole job — one deliverable may contain both AI-assisted and primarily human steps.

8. PROVE: Client Task & Information Plan CAPSTONE EVIDENCE

Create a plan covering the full BrightPath capstone. Keep this — it becomes part of your capstone evidence in Lesson 7.3.

9. Human Judgment Checkpoint

Before moving to 7.2, ask: could I explain Sarah's assignment without rereading the entire handoff? Do I know which source materials belong to each deliverable? Have I identified information I should not casually give to AI? Do I know what still requires clarification or human judgment? If not, stop and clarify before producing the work.

10. Reflection

11. Lesson Close

Good client work starts before the first draft. You need to understand the assignment before asking AI to help with it. In the next lesson, you'll use this plan to produce the actual client operations package.

Build the AI-Assisted Client Operations Package

This is a multi-session capstone production lesson — you'll build five connected deliverables for BrightPath Learning Co., divided into checkpoints so you can complete, review, and organize each piece before moving to the next.

Checkpoint 7.2a — Client Welcome / Introduction Email

Task: Use the BrightPath client information and communication preferences to produce a concise, client-appropriate introduction or welcome email.

Draws on Lessons 2.2, 2.3, 3.1, 3.3, 3.4. Source: Brand & Communication Guide and Sarah's Email, item 1.

Review before moving on:

• Purpose
• Accuracy
• Tone
• Client voice
• Audience fit
• No unnecessary AI-generated wording
• Professionalism

Checkpoint 7.2b — Customer Service Response

Task: Handle a customer request using approved BrightPath information. If the available information doesn't support an answer, handle the uncertainty appropriately rather than inventing a policy.

Draws on Lessons 2.2, 3.2, 3.3, 3.5. Source: Customer Support Thread.

Review before moving on:

• Approved information only
• No unsupported claims
• Customer tone
• Clarity
• Escalation requirements
• Final accuracy

Checkpoint 7.2c — SOP

Task: Turn the supplied BrightPath process information into a usable SOP. Do not invent procedures simply because they seem reasonable — if an important step is missing from the source, identify it rather than silently creating a new process.

Draws on Lessons 2.2, 2.3, 5.1, 5.2, 5.3. Source: Existing Onboarding Confirmation Process and Meeting Transcript.

Review before moving on:

• Source information
• Sequence
• Missing steps flagged
• No invented procedures
• Exceptions
• Completion check
• Usability

Checkpoint 7.2d — Weekly Status Report + Supporting Spreadsheet/Data Work

Task: Prepare the BrightPath weekly status report using the relevant work and business data. The supporting spreadsheet work must contribute to the report — don't treat it as unrelated practice. Recreate or organize the supplied business data in a spreadsheet, then use that work to support the status report — the handoff pack gives you the data as a table to work from, not a downloadable file.

Draws on Lessons 3.6, 5.4, 5.5, 5.6. Source: Sarah's Email, item 4 and Business Data. Report structure: Completed / In Progress / Needs Attention / Next Steps.

The spreadsheet supports the status report and its evidence — it is not required to become a separate portfolio artifact. Review before moving on:

• Status accuracy
• Calculations
• Formulas
• Data organization
• Supported observations, no unsupported conclusions
• Report clarity
• Spreadsheet actually supports the report

Checkpoint 7.2e — Research Brief

Task: Complete the assigned BrightPath research task and produce a concise, verified research brief.

Draws on Lessons 4.1–4.5. Source: Research Request & Prior Notes — note that Daniel's draft summary line is explicitly flagged as possibly unsupported; that's the verification skill from Lesson 4.4 in action.

Review before moving on:

• Source quality
• Factual accuracy
• Verification
• Findings vs. implications distinguished
• No unsupported claims
• Limitations stated
• Client usefulness

Capstone Production Rules

1. Keep the source material — you'll need it during the audit stage. 2. Keep the AI-assisted work — where AI contributed, keep the relevant draft to show how the final work developed. 3. You don't need AI for every deliverable — if a task is faster, safer, or more appropriate without AI, complete it manually. 4. Review before moving forward — don't wait until the end to discover source material was misunderstood. 5. Keep your work organized — maintain the relationship between the original task, source material, AI-assisted work, audit, changes, and final deliverable. This makes Lesson 7.3 much easier.

Capstone Evidence Sequence

The purpose is not to prove that AI did the work — it's to prove that you supervised the work responsibly.

Human Judgment Checkpoint

Before finishing the production stage: what decisions did I make that AI could not make for me? Consider: what information to use, what not to use, what to ask AI to do, what to verify, what to reject, what to change, whether the final result was actually ready.

Lesson Close

You've now moved from individual AI exercises to connected client work. The next stage is not to produce more — it's to audit, correct, and finalize what you've produced.

Audit, Correct and Finalize

1. Why This Matters to a VA

Producing a draft is only part of the job. A client doesn't receive "here is what AI generated" — they receive "here is the finished work." A strong final review is not simply proofreading. It asks: is this correct? Does it answer the client's actual request? Is anything missing? Did AI add anything unsupported? Did I accidentally change the meaning? Is the communication appropriate? Did I handle information appropriately? Can the client use this without having to fix it?

2. The Audit Comes From the Task

Don't audit every deliverable in exactly the same way. Start with the original client request — the Client Handoff Pack is the authoritative source for this audit: if a deliverable claims something the pack doesn't support, that's an error to catch, not something to assume is fine. Keep your Client Task & Information Plan from Lesson 7.1 as capstone evidence too, since it shows how you understood the assignment before production began. Then identify the checks that matter for that particular deliverable:

3. Full AI Audit Note

Complete one full audit note here in the course, for your strongest deliverable. For the other four, complete the same 12-field structure in your own external capstone working file (a doc, not this page) — they're still required capstone evidence, just not captured in this form, since five full audit notes here would make this page unwieldy to actually use.

4. Demonstration: Audit Before and After

Take one capstone deliverable and compare the original AI-assisted version with the final corrected version. Identify: what AI produced, what was wrong or incomplete, what you changed, why the change was necessary, and how you verified the final version. The important evidence isn't that the second version looks better — it's that you can explain why it's better.

5. Deliberate Risk

One capstone deliverable contains an error that's easy to overlook — a status changed from pending to completed, a research finding presented without adequate support, a spreadsheet calculation using the wrong range, an SOP with an invented step, a customer response promising something the source doesn't confirm, an email with an incorrect date or time, or a statement including information that shouldn't have been used with AI.

6. Student Practice: Audit All Major Deliverables

For each: return to the original task, identify the important requirements, identify the risks that matter, inspect the final work, compare it against the source material, correct problems, verify the corrections, and complete the Full AI Audit Note. Don't make changes simply because a sentence could be written differently — a change should have a reason.

7. Human Judgment Checkpoint

For each major deliverable, identify at least one important decision that required your judgment — e.g. "I rejected an AI-generated claim because the source didn't support it," or "I corrected the spreadsheet because the suggested formula used the wrong range."

8. Client-Ready Review

• Accuracy — is it correct based on the available evidence?
• Completeness — did I address the actual request and required information?
• Client Fit — does it match the client's expectations, audience, and communication style?
• Verification — did I verify information that required verification?
• Privacy — did I handle client and customer information appropriately?
• AI Review — did I check for AI-generated errors, assumptions, or unsupported claims?
• Professionalism — would I be comfortable sending this to the client?
• Usability — can the client actually use this deliverable?

9. PROVE: Final Audited Client Operations Package CAPSTONE EVIDENCE

The package should contain the corrected final versions of all five checkpoint deliverables plus the required Full AI Audit Notes for the major deliverables.

10. Reflection

11. Lesson Close

The final review is part of the work. AI can help produce the draft. The VA decides whether the work is ready to deliver.

Portfolio Evidence and AI-Use Explanation

1. Why This Matters

A portfolio shouldn't simply show polished documents. A client may want to know: can you understand instructions? Work with messy information? Use AI without blindly trusting it? Catch errors? Protect client information? Communicate professionally? Explain your decisions? Deliver useful work? Your capstone gives you evidence for those questions. The goal isn't to show everything you created — it's to select strong evidence that demonstrates capability.

2. The Evidence Sequence

This sequence shows the work as a process — it demonstrates the final result didn't simply appear from an AI tool.

3. What Makes Evidence Strong?

Strong evidence shows more than a finished document — it shows Context (what was the client asking for?), Work (what did you actually produce?), AI Contribution (where did AI help?), Human Contribution (what did you decide, verify, edit, or reject?), Quality Control (what problems did you catch?), and Final Result (what did the client ultimately receive?).

4. Selecting Evidence

You don't need every exercise from the course — prioritize work that demonstrates different capabilities: Communication (Welcome Email), Customer Support (Customer Response), Operations (SOP), Reporting and Data (Status Report + Spreadsheet), Research (Research Brief), and Process Evidence (selected Audit Notes showing meaningful human judgment). The goal is fewer, stronger, more varied pieces of evidence.

5. AI-Use Explanation CAPSTONE EVIDENCE

Prepare a concise explanation that answers each of these:

6. PRACTICE: Explain One Deliverable LEARNING ARTIFACT

Choose one capstone deliverable and explain it in a short client-facing format: what the task was, how AI helped, what you checked, what you changed, what remained your responsibility. Don't describe AI as if it independently completed the client assignment.

7. PRACTICE: Client Question Practice

Imagine a client asks: "Did you use AI to help with this?" Practice answering honestly and professionally — acknowledge AI use when it occurred, explain what it helped with, explain that you reviewed and verified the output, avoid claiming AI did work it didn't do, avoid hiding AI use when disclosure is required, and follow the client's AI-use policy when one exists. The correct response depends on the actual work and the client's expectations.

8. Human Judgment Checkpoint

Complete this sentence: "The most important part of this project that required my judgment was..." Then explain why AI could not simply make that decision for you.

9. Portfolio Case Study Structure

1. Client / Scenario — briefly introduce BrightPath Learning Co. and the type of VA work involved
2. Client Need — what problem or assignment needed to be handled?
3. Work Completed — summarize the major deliverables
4. AI-Assisted Workflow — explain where AI was used
5. Verification and Quality Control — show how you checked the work
6. Human Corrections — specific examples of changes you made
7. Final Deliverables — show selected final work
8. What I Learned — how your approach to AI-assisted VA work changed

10. What Not to Include

Avoid turning the portfolio into a collection of random AI screenshots, a list of prompts, a list of AI features, unsupported productivity claims, exaggerated AI-expertise claims, every exercise from the course, or raw AI output presented as finished client work. The portfolio should demonstrate VA capability, with AI as part of the workflow.

11. PROVE: Portfolio-Ready AI-Assisted VA Case Study CAPSTONE EVIDENCE

Your final case study should include: selected client task, relevant source/context, AI-assisted work where appropriate, human audit, important corrections, final deliverable, AI Audit Note, and a concise AI-use explanation.

12. Final Course Reflection

You've reached the end of the course. Take a moment with these before you close it out.

13. Lesson Close

Your portfolio does not need to prove that you can make AI produce impressive output. It needs to prove that you can handle real client work with AI responsibly. That means understanding the task, using AI where it helps, checking what it produces, making the necessary corrections, and taking responsibility for the final work.