Phase 1 — Data Input & Gathering

Welcome to Module 1, where we start the first of the 8 Work Phases for White-Collar Professionals (WCPs). This phase is Data Input & Gathering. You’ll see how to bring all your relevant information (internal or external) into an AI-ready format, so later phases can build on these inputs.

We’ll follow a single unfolding scenario at our hypothetical company Organic Fruits Global. focusing on multiple departments (Marketing, Operations, Finance, etc.) that must gather data differently but ultimately feed it into one cohesive AI-driven workflow.

Lesson 1: Data input basics and methods

1. Why Data Input & Gathering Matters

At Global Solutions Inc., each department deals with diverse data:

• Operations collect production logs and quality check forms (sometimes handwritten).

• Marketing obtains social-media analytics, lead-generation forms, and competitor info from web searches.

• Finance receives invoices, vendor contracts (often PDFs), and meeting recordings discussing budget changes.

The input is any information you provide to the AI. Information comes in all forms and all of them matter. Each data inputting technique is specifically curated to get your desired result faster or more accurate and the recent updates in image, video and real time feed have opened a ton of new possibilities for productivity increases.

Key reminder: None Of these should be used except with the 3 main pillars of the 6 pillar prompt. (Task, Context & Specifications) they only aid you in getting and even better answer in less time.

2. Methods of Data Input

A. Text Input

1. Direct Typing / Copy-Paste

   • Use this for any plain text (emails, articles, short notes).

   • Marketing might paste competitor product descriptions into ChatGPT to see how they compare.

2. Importing/Attaching Files

   • If you have short PDF or Word files, you are able to attach them directly in ChatGPT for a quick summary.

   • All common file extensions for text files, spreadsheets, presentations, and documents can be inserted

Scenario Example

Marketing wants to share a product launch text file with ChatGPT to draft more compelling ad copy. They copy-paste their existing description, they attach a PDF containing all the products info for extra context and provide chatgpt with a persona,specifications for output and examples of how the ad copy for other products lookedlike

B. Voice Input

1. Mobile, desktop and OS apps

   1. Tap the mic icon and start speaking

   2. tap the mic and start talking

C. Image Inputs (OCR) & Real-Time Vision

1. OCR (Optical Character Recognition)

   • You can input any image into the Chat Which:

     • Saves Time

     • Increases accuracy (by providing visual context)

   Examples:

   • it also saves alot of time

   Example:

   • Finance might scan vendor invoices or receipts for quick data entry.

   • Marketing Might insert their Instagram post for design feedback

   • Product Managers can ask for feedback on products package design

2. Screenshotting:

   1. Why: Screenshot your work instead of talking about it reducing the amount of context you need to provide

   2. How

      1. Click the Screen shot icon on your ChatGPT app

      2. Screen shot your current screen

3. Real-Time Vision and voice mode

   Reminder: for the Real-time vision its still better to provide context and specifications at first and then as you go through the meeting you can give it new task prompt by voice

   • Using the Advanced voice mode you can have ChatGPT capture input in real time which is extremely useful for Non-deep work such as:

     • Quick Work

       • Meetings: when in a meeting having advanced voice mode present hearing you while you convers is gonna allow you to:

         • Ask its opinion on presentation as they are presented

         • Have it search for something in the meeting that you need to know without disturbing the meetings Flow

         • and much much more its the most useful feature for immediate and quick tasks

     • Deep Work:

       • Uninterrupted: Leave it on, on your phone app and put your phone on its face. Now while your working you can ask it questions without interrupting your workflow

       • Analysis: You can use the Analysis techniques in module 3 without interruption or physical typing

Scenario Example

HR has a sign-up sheet from a job fair with 40 potential candidates’ names. They snap a photo, feed it to ChatGPT and instantly convert it to a neat CSV list and start conversing about it in the meeting and analyzing the data (more on data analysis in module 3)

Lesson 2: MP4/MP3 Inputs & Gemini 2.0 Screen Sharing

1. Gemini 2.0 Flash

   • Google’s advanced AI system that handles real-time camera feeds and screen shares.

   • You can

     • Share your screen in real time with it

     • Insert a video or audio

   • You will use Gemini for now by either using the same prompting techniques you were applying in ChatGPT in Gemini or get ChatGPT to ask questions to Gemini by explaining to ChatGPT that ” the video or audio your working in is in the hands of another AI so if you need anything from me ChatGPT please ask”

   • When openAI releases the Same feature we will add it to the course instead of Gemini 2.0 for now use Gemini if you deal with video and audio a lot and would benefit form Gemini

Scenario Example

Finance hosts a remote budgeting meeting, shares the screen with Gemini 2.0. It flags lines in the Excel doc where cost overruns appear. Those flagged data points are then compiled for Phase 2 (Data Organization).

c

4. Assignment: Practice Data Input

1. Choose One work Scenario

   • Choose an example that's common for your work and preferably have data reports and files on it because you are gonna build on that example for each module

2. Gather Data

   • Use at least two methods of the data inputting method plus attach a data report or any sort of data that you can make a decision on: e.g., voice input for notes + web search (DeepSearch or Perplexity) for external facts.

3. Ask ChatGPT to Summarize

   • Paste everything into a ChatGPT conversation or a single “Project”

   • wait for module 2 instructions

4. Reflect

   • Did you skip any method that might help?

   • what could you have done first or used a different method that could have made the process faster?

   • take a second and think in your opinion which method would be the most used method for your type of work?

Key Takeaways & Next Step

Phase 1: Data Input & Gathering is about combining all relevant info—internal logs, external research, scanned documents, voice notes, or live video—and channeling it into an efficient workflow.

• No single method is “the best”; pick what fits your common tasks and the data you most commonly use.

• Next, in Phase 2 , you’ll see how to organize this raw data, efficiently working through it, building structured formats, and preparing it for deeper analysis.

Congratulations on finishing Phase 1! Let’s move on to Phase 2—Data Organization where your newly gathered info turns into well-structured data for advanced AI workflows.

Phase 2 — Data Organization using the book framework

Welcome to Phase 2 of our White-Collar AI Workflow! In Phase 1, you gathered all sorts of data (text files, voice notes, images, external research). Now it’s time to organize that raw information so future steps—like analysis, decision-making, and reporting—run smoothly.

Think of Phase 2 as cleaning and structuring data. Whether you have PDFs, meeting transcripts, or external stats from DeepSearch, this phase turns that jumbled input into neat, labeled, or summarized pieces.

Lesson 1: The book Framework

1. Why Data Organization Matters

• Clarity: Organized data prevents confusion when multiple departments (Finance, Operations, Marketing) reference the same info and especially when a lot of data has to be taken into consideration in order to make a corporal decision.

2. The Book framework

A. Structuring

1. Clarify base

   1. Ask ChatGPT “What do you understand and what can you immediately notice in the data”

      1. this sets the base for ChatGPT and your own understanding of what your going to do

         1. Merging External Facts

            • If ChatGPT’s internal knowledge misses the latest updates, you can copy fresh info from Perplexity (like an industry trend discovered last week or certain public data reports) and feed it into ChatGPT.

            • Example: Marketing found a brand-new competitor ad campaign (via Perplexity). They paste that snippet into ChatGPT, asking: “Compare this new campaign to our Q3 approach. Spot any weaknesses.” allowing an analysis from another external angle

            • Note: you can streamline this for searches that aren't too complicated by just having ChatGPT do the web searching itself using the /websearch command but perplexity is generally better at search and deepsearch even better for complex deep research

         2. Multi-Layer Queries

            • If your data is big, break it into chunks: “First, ChatGPT, identify the biggest cost items in each invoice set. Apply the refinement cycle technique that your gonna learn on it then move on to the next item

         3. Handling Complex Scenarios

            Real business data isn’t always straightforward:

            • Missing or Contradictory Info: If one department’s data conflicts with another’s, ask ChatGPT to highlight or propose reasons.

            • Sensitive/Confidential: If certain data is private or if you want to keep it out of ChatGPT, you can create synthetic placeholders or partial data that align with our own data trends. Then see if ChatGPT can still glean patterns from what remains. (But this should only be used on linear data that you are sure is consistent)

   2. Now Give ChatGPT your full 6 pillar prompt

      Pro Tip: Use the interrogation pattern

      1. What other data do you need to confirm or disprove this hypothesis

      2. You can instruct ChatGPT: “If you see contradictory numbers, show me possible reasons they might differ.” This spares you from hunting discrepancies manually.

2. Create an Outline or Index

   • If you have multiple files or transcripts, you can build an outline for each.

   • Example: After you give ChatGPT the info.

     • in your task prompt tell it that “

       • Template

         for now you should organize the info i gave you according to the context and the task we are gonna be working on today in a table of contents or outline that would enable us to take this process step by step without forgetting or confusing anything”

3. Now work on this outline step by step by saying for example

   • Template

     Okay so lets star with part 1 about wither finance is responsible for the lower profit margins we are experiencing could you outline what data finance has given us.

     Next

     Okay so lets start with analyzing Fixed costs report for software X that we have been using lately and see XYZ

3. Workflow Example (Global Organic Fruits)

A Website Sales Specialist has been assigned a task by the Chief marketing officer to increase the conversion rate (which is the percentage of people who take the action that you desire them to take on that page) of the landing page of the business website

1. Gather: he then gathers all the data that directly or indirectly affects that conversion percentage

   • the ad that sends them to that page

   • the Data reports on stuff relevant to his job

     • Heat maps

     • scroll map

     • Time spent on page

     • speed of that page

   • Customer support and Customer success agents reports which might have insights on what customers encounter an error with.

2. Input:

   1. He inputs those data files into ChatGPT

   2. He provides ChatGPT with this prompt

      You are a senior conversion rate optimization specialist with over 30 years of experience

      Task< you are gonna find out what test i am going to conduct on this landing page that could yield me the highest ROI on my time spent on task based on the data provided

      Context< we are an E-commerce business that sells organic fruit through our website and we market to over 40 year old mothers who are concerned about their long term health and processed chemicals in the attached files your gonna find a script for the exact Ad copy we are using to target these ladies we market through google ad search. i am a CRO specialist working for this company and i conduct regular a/b testing on the landing page to see which version would get me better conversion percentages. you will find reports that greatly specify what our users do after they enter the page>

      Specification for your output< your output should be: 1) very organized 2) very logical and professional 3) concise but not missing any important info that might aid in figuring out the best test to conduct>

      Specifications for input< we are gonna take this step by step that's why first i want you to create a step by step outline based on the data i provided and the context i have given you that we can follow to reach our desired conclusion>

      here are some examples of a previous chat conclusions that we did last week< Paste example>

      Notes< remember to not forget anything i told you in the middle of the context as all of these aspects combine together to form a big picture for you so you can better understand my customer’s journey>

4. Assignment: Data Organization Exercise

1. Now organize your data by writing a 6 pillar prompt

2. After creating your outline wait for instructions in module 3

5. Key Takeaways & Next Step

• Data Organization is about cleaning and structuring your Phase 1 inputs so you can assign more computing power to each of the phases in finding your desired outcome as AI’s still cant do such large jobs in a single prompt or without good guidance and it also helps you organize your workflow to prevent human error

• Detailed Techniques: Merging external research, chunking large data, and side-by-side comparisons help ChatGPT provide a more accurate Analysis.

Coming Up: Phase 3 — Data Analysis

With everything tidy, you’ll see how ChatGPT can sift through your organized data, refine it, and help you find real insights. Stay tuned for Phase 3!

Phase 3 — Data Analysis using the refinment cycle

In Phase 2, you turned messy input data into neat, well-labeled pieces. Now, in Phase 3, we’ll use that organized data to discover insights, spot patterns, and plan next steps at Global Solutions Inc. This is where AI tools really shine—finding hidden trends or answering tough questions that might take a human team days or weeks to figure out.

Lesson 1 : refinement cycle framework

• Deeper Insights: Using AI is not only extremely faster but even provides better quality or insights you couldn't have thought of in a million years because it can see all the data all at once at the same time from all the different business angles.

• Faster Problem-Solving: Instead of endless manual calculations, ChatGPT can quickly highlight patterns, anomalies, provide solutions & assessments

• Cross-Department Value: Finance, Marketing, and Operations all benefit. For example, a single transcript from a budget meeting plus competitor data can reveal cost-saving strategies that ChatGPT can immediately think of ways to apply it to each department.

Scenario at Global Organic Fruits.

• Operations has shift logs and production data from Phase 1.

• Marketing gathered competitor info, web traffic stats, and social media leads.

• Finance has updated invoices and meeting notes about cost allocations. In Phase 2, they organized it. Now, in Phase 3, they feed it to ChatGPT for real analysis—looking for production bottlenecks, comparing marketing ROI, or verifying budget solutions.

2. Refinement Cycle Technique

What: A process of iterative questioning and improvement—ChatGPT or your AI tool proposes a hypothesis, you use the AI to identify gaps and problems in the hypothesis provided and ask it to refine it again and repeat the cycle leading to an almost perfect solution.

Steps

1. Outline Goal through a follow up prompt: “Analyze these 3 competitor reports and these 2 internal sales logs to find any potential improvements we can make or mistakes to avoid to achieve our desired outcome”

2. Get Initial Analysis: ChatGPT might highlight a potential gap in:

   1. how many people click on the ad that actually make it to the landing page

   2. industry average cost of production of competitors in China versus our own cost of production highlighting a potential issue to be solved there

   3. highlight to an operations manager the inefficient use of customer service employees at peak call times could be better

   • Template

     "We’re gonna analyze the data to identify potential improvements, solutions, or issues using [data, reports, or documents]. Here’s the initial task: Analyze [insert data/reports] to identify potential improvements, solutions, or mistakes to avoid in achieving our desired outcome of [state desired outcome]. Follow these steps: Provide an initial analysis based on the input data. Identify gaps, patterns, or potential issues worth exploring further.

3. Self-Criticism: Ask ChatGPT to critique its own result: “List potential errors in your analysis or areas you might have missed.”

   • Template

     Self-critique your analysis by identifying any potential errors, omissions, or limitations in the conclusions provided.

4. Refine: Tweak the instructions or add missing details. Let ChatGPT run the analysis again.

5. Repeat until you see no major flaws or discover every angle worth exploring.

6. Choose the top 3 or 5 hypothesis in order to make a decision in the next phase

7. Now you have every potential hypothesis on what the issue or solution is alongside their supporting or refuting arguments

Why It Helps

• You can tell ChatGPT: “Look at this from an Operations angle too,” or “Check if we missed mention of competitor Y.” allowing you to capture angles outside your own professional scope.

• It also disproves itself by telling you why it could be wrong from angles you couldn't have possibly considered on your own leaving you with only the decision to make (in phase 4 we are gonna understand how to make the decision based on the analysis conducted)

5. Workflow Example

1. Operations

   • They feed last quarter’s organized shift logs + production outputs.

   • ChatGPT identifies a pattern: certain lines slow after 2 p.m. on Mondays.

   • Self-criticism step reveals ChatGPT can’t see external weather data. Operations manager fetches that info from Perplexity, suspecting humidity issues. ChatGPT merges these insights and confirms high humidity correlates with slower output.

2. Marketing

   • Marketers combine competitor social media stats + their own site analytics.

   • ChatGPT sees that competitor runs daily LinkedIn polls that drive strong B2B engagement.

   • The team re-checks the data to confirm, then tasks ChatGPT with a short bullet plan to replicate that success.

3. Finance

   • They input an updated invoice dataset (organized in Phase 2).

   • ChatGPT flags a vendor with rising costs, which might be negotiated down.

   • In the refinement cycle, finance gives more vendor context (“We switched suppliers last year”), and ChatGPT clarifies potential renegotiation strategies or alternative suppliers found via DeepSearch.

Now lets continue our own example by analyzing the data we provided for our E-Commerce organic Fruit company

6. Assignment: Data Analysis Exercise

1. Continue on the same chat you used in phase 3

   1. Choose one or more of the analysis frameworks we explained and generate multiple hypothesis to reach your desired outcome and generate the hypothesis that disproves them

   2. Remember:

      1. Ask it to view the issue from different angles that you might not be considering

   3. Choose the top 5 or 3 hypothesis that you think are actually true (We are gonna learn how to make a result and decide in phases 4,5)

Key Takeaways & Next Step

• Phase 3: Data Analysis relies on well-organized inputs from Phase 2.

• Refinement & Self-Criticism ensures you don’t just accept the first AI result, but instead use the AI to criticize its own analysis making the analysis process not just much more accurate and higher in quality but also saves up time.

Coming Up: Phase 4 — Data Confirmation

After analyzing your data thoroughly, you’ll learn how to make decisions that yield you the highest ROI on your time and resources.

Phase 4 — Confirming Data & Results

In Phase 3, you analyzed the data and created 3 to 5 core hypotheses.
Now you must verify everything before making a decision.

Lesson 1: Internal & External Data Verification

1. Why Confirming Data & Results Matters

Accuracy:
Even strong plans fail if based on incorrect data

Credibility:
Verified data builds trust with your team

Risk Reduction:
Unverified data leads to poor decisions

Scenario:

• Marketing launching a campaign → Is competitor data current?

• Finance negotiating costs → Are estimates accurate?

2. Internal vs External Data Verification

Once your logic is clear, verify the data used.

Internal Data:

• Recheck with internal systems or teams

• Confirm numbers match original data

External Data:

• Confirm information is up to date

• Use reliable and trusted sources

• Avoid unknown or low-quality sources

Using ChatGPT for Verification

• Ask it to list the sources used in your report

• Ask it to check for inconsistencies

• Ask it to act as a critic and find weaknesses

Optional: Peer Review

If data came from a colleague:

Use ChatGPT to summarize your conclusions and actions clearly, then send it for confirmation.

Scenario 1:

Finance reviews shipping cost data.
They confirm the data is current before using it in negotiations.

Scenario 2:

Marketing reviews competitor ad data.
If data is outdated, they flag it and adjust their conclusions.

5. Workflow Example

Operations:

• Confirms data with internal logs

• Ensures no missing information

Marketing:

• Verifies competitor data

• Updates any outdated insights

Finance:

• Confirms cost data with internal records

• Checks all calculations before approval

6. Key Takeaways

• Always verify before deciding

• Check both internal and external data

• Use reliable sources only

• Fix gaps before finalizing decisions

Next Step:

Phase 5 — Decision Making

Phase 5 — Decision Making

In Phase 3 we have analyzed all the potential hypothesis that we can use to achieve our desired outcome. in phase 5 we are then gonna decide which hypothesis is the most accurate and can provide us with the highest ROI on our time and resources

Lesson 1: Predictive Scenario Modelling framework

1. Gathering Team Inputs & Voice Q&A

Real decisions aren’t made alone. You might need group input from managers or staff if that's the case then:

1. Collect needed input & opinions to make a valid business decision

   • In a meeting, each department can “ask” ChatGPT (through voice mode or typed prompts) clarifying questions.

   • Or when you filter the hypothesis’s down to the top 5 or 3 it would be clear which departments you need their inputs on the analysis ask them if need be then return to ChatGPT with their inputs to avoid making a useless decision

   • Example: “Finance Manager says: ‘What if the budget shrinks by 10% next quarter?’ ChatGPT then recalculates predictions or narrows the scope.”

   • Example: If ChatGPT suggests cutting line #3 staff, but HR points out morale risks, you add that perspective to finalize the decision.

Practical Tip

If you record a Zoom or in-person meeting with a tool like Otter AI or FireFlies, you can transcribe everyone’s feedback. Then feed the transcript into ChatGPT, asking: “(Summarize/take into consideration) the top arguments from each department regarding Plan A vs. Plan B.”

Interrogation pattern

if your unsure about which departments have a say in this for any reason ask ChatGPT once again if it needs any particular departments input or opinion. Don't forget to explain to ChatGPT what your role is in all of this.

2. Predictive Scenario Modeling

1. Playing Out Hypothetical Scenarios:

   • This is our Exclusive technique that we use to explore different hypothetical scenarios from different business angles in order to have the full view picture of what could happen based on the extensive data and analysis you have conducted

2. Steps:

   • Give ChatGPT everything you did so far and how it led to the top conclusion or hypothesis that you have right now

   • Ask ChatGPT to clarify what it understands so far and ensure that its understanding aligns with yours

     • Template

   • Ask for thoughts on the different paths and provide more context if needed by having ChatGPT ask you questions about what it needs (interrogation pattern)

   • now Ask ChatGPT to play out the different scenarios in great detail. but take it scenario by scenario

   • Take your time to play out the scenarios extensively and even ask ChatGPT to look at it from a different perspective (e.g: customer perspective, business brand perspective, long Vs short timeline perspectives)

   • Ask for what ChatGPT thinks the right decision should be

     • if you're not satisfied with the answer then provide ChatGPT with your thoughts

     • work them out (Cycle of refinement)

     • then ask it its opinion again

     • keep repeating till your are satisfied with the answer

   • Now you have a final result of what you should do next or others should do

     • Template

       "I need to play out hypothetical scenarios for [specific situation/project/decision] to gain a comprehensive understanding of potential outcomes. Remember what we have done in this chat so far/Here’s what I’ve done so far:

       • [Describe everything you’ve done, key data, and analysis that led to your current hypothesis or conclusion if its not already in the chat].

       First, summarize what you understand so far based on this input. Ensure your understanding aligns with mine. If any clarification is needed, ask me targeted questions (interrogation pattern) to refine your understanding.

       Next, take the following steps:

       1. Play out potential scenarios, one at a time, providing detailed outcomes and reasoning for each scenario.

       2. Evaluate these scenarios from different perspectives, such as [e.g., customer perspective, business brand perspective, short-term vs. long-term outcomes, financial impact, etc.].

       3. Based on the explored scenarios, recommend the best course of action with reasoning.

       If I’m not satisfied with your recommendation, I will provide my thoughts, and you will refine your suggestions accordingly. Continue the cycle of refinement until we reach a satisfactory conclusion.

       Please make the analysis detailed, logical, and easy to follow at every step."

Scenario Example

An HR Recruiting Employee after Analyzing the different candidates for a Product Design Role decides on which one of the top 5 Candidates would be best for the role based on multiple factors such as culture alignment, metrics that candidate knows, the candidates personality and work ethic and similar roles worked by the candidate and so much more that the HR employee might have not considered or may have taken a long time to go through them and through what could go wrong.

Pro Tip

Whenever ChatGPT suggests an action plan, consider running the Refinement loop again. For instance, “Which steps might cost too much? Give me a cheaper variant.”

Pro Tip

After ChatGPT can no longer identify potential issues, ask it to look at the idea from different perspectives (e.g., sales, legal, product fulfillment) to ensure all potential problems are covered.

4. Finalizing Decisions & Minimizing Risk

Once you’ve weighed the outcomes and considered each department’s perspective:

1. Write your Final Decision

   • ChatGPT can produce a detailed or short summary:

     • Template

       “Now that we have done an extensive analysis with arguments for each decision both for and against the decision and we have decided on decision X i want you to write in an extremely linearly organized manner what are the steps we took and outcomes we weighed in order to choose this decision and why we chose this decision and what we decided to do on the risks and downsides associated with this decision in order to mitigate them. this extensive detail will serve as a basis for presentation and Supervisor review for my company.

3. Workflow Example (Global Solutions Inc.)

Continue to make a decision using the predictive scenario technique on the top 5 hypothesis you chose in phase 3

4. Assignment: Generating Results Exercise

1. Continue on your choice of example from phase 3 and 4

2. Now Use the predictive scenario technique to make a decision

3. After you have made the decision ask ChatGPT “provide me with an extensively detailed summary of the steps i used to make this decision that i will provide to a teacher who is teaching me how to use techniques that help me in my job such as the ones i used in this chat”

4. Paste the answer in the comments for peer review

5. Key Takeaways & Next Step

• By using the predicative scenario technique you have:

  • Conducted a thorough analysis that encompasses multiple aspects to determine the highest ROI path with the lowest Risk

  • Have clearly laid out your decision making process and what its based upon for any supervisor review or your own future reviews on the outcome of this decision

Phase 6 — Communicating Results

You’ve made a decision in Phase 5 (Decision Making). Now, in Phase 6, it’s time to effectively communicate the outcomes. That could mean presenting data insights to your boss or sending instructions to your team. Whether you’re delegating tasks, visualizing data, or crafting a presentation, Communicating Results ensures everyone understands what’s happening and how to move forward.

Lesson 1: 101 Communication

1. Why Communicating Results Matters

Did you know?

Miscommunication is one of the most common problems that businesses face

• Clarity: Even the best plan fails if it’s poorly communicated.

• Team Alignment: Finance, Marketing, and Operations each get the same message tailored to them and what they should do next—reducing miscommunication thus enhancing efficiency

• Actionable Follow-Through: Clear instructions or data stories help people take the next steps quickly.

1. Delegation Technique

• Use any of the data organization methods to organize the data to the point of clarity.

  • Side Note: The degree of finalization in the email or instruction you send is based completely on your liking. If you prefer to analyze the data and then send it for others to conclude results—or conclude results yourself and then send it for communication (e.g., to the creative design or sales team)—it is completely up to your processes.

• Explain the task that the employee receiving the email will perform, and provide necessary context (e.g., their role, name, responsibilities).

• Ask ChatGPT to “break down the task into the simplest, easiest, and most broken-down instructions that even 5th graders can follow but address them in a tone that suits a white-collar professional.”

• Ask ChatGPT if it thinks anything is missing or should be added/removed to ensure clarity and avoid miscommunication.

• Review and refine the instructions to ensure clarity.

  • Instructions can be customized to your preferences (e.g., explain why every step is asked of this employee, instruct them to respond promptly, provide a time estimate or deadline, etc.).

• Template

  "I need to delegate a task to [employee’s role, e.g., a creative designer]. The task is to [describe task, e.g., create a presentation, analyze sales data]. Write a professional email that includes:

  • A brief introduction and context explaining the task's importance.

  • A clear breakdown of steps they need to follow to complete the task.

  • A request for them to use the provided ChatGPT link for further clarification.

  • Clear deliverables and a deadline.

  • A tone suitable for a white-collar professional but simple enough to avoid ambiguity.

  • Encourage them to reach out promptly if they need additional assistance or have questions."

2. Sending Instructions

• Send the email with the ChatGPT chat link for better context and to reduce the time spent explaining the task.

  • This allows the instructed employee to enter the same chat and ask ChatGPT questions like:

    • "What does this mean?"

    • "Should I do this?"

    • "What is this based upon?"

3. Optional Step

• You can also have ChatGPT translate the instructions into another language entirely for remote workers.

Lesson 2: Creating Image Visualizations with ChatGPT

1. Introduction to Data Visualization

• The process of representing data in Visual formats to make it easier to understand.

  • (Note: There is AI art, but we are focusing purely on visualization from a work perspective.)

2. Types of Visualizations

• Files: PDF, CSV, DOC, XLSX, PPTX, ZIP.

• Charts: Bar charts, line charts, pie charts.

• Graphs: Scatter plots, histograms, heatmaps.

• Tables: Structured data representation.

• Basically Everything you could need

1. Steps to Create Image Visualizations

• Specify exactly what you want to visualize, For Example:

  • Example 1: A bar graph demonstrating the relationship between age and the likelihood of purchasing.

  • Example 2: A plot graph showing the relationship between customers' subscription dates and their average LTV.

  • Example 3: A representation of your choice that shows customer satisfaction percentage growth over the past 10 years (2016 to 2026) and predicted growth for the next 10 years (2026 to 2036).

    • Use these specifications:

      • Line color for future predictions: Pink.

      • Line color for past representations: Blue.

      • Dates: Bold Arial font, Dark Black, font size 16.

      • Background: #FAF9F6.

      • Y-axis (Customer satisfaction percentage growth): Turn pink midway where it shows future predictions.

      • X-axis: Years.

• Specify the type of visualization (e.g., bar chart, pie chart).

• Ask ChatGPT to generate the visualization.

2. Pro Tips

• Ensure data is clean and well-organized before visualization by following previous phases.

• Use clear and detailed specification prompts to get accurate visualizations.

Lesson 3: Storytelling AKA Verbal Communication

1. Use Case

• Using ChatGPT to create engaging data stories that communicate complex data insights in a simple and engaging way, making them easier to understand and follow.

2. Steps

1. Prepare the Data:

   • Collect and organize relevant data for the story. Ensure the data is clean, accurate, and complete.

2. Define the Story Structure:

   • Determine the narrative structure for the story, or ask ChatGPT to define it.

   • Focus on logical flow and clarity.

3. Generate the Story:

   • Use ChatGPT to generate the story based on the prepared data and structure.

   • Ensure the story clearly and linearly presents the data and explains what conclusions are drawn, along with the reasoning.

   • Template

     "Based on the provided data, define a logical narrative structure for a story that will explain these insights to [specific audience, e.g., executives, team members, clients]. The structure should include:

     • Introduction: Context and purpose of the story.

     • Body: Key insights, data trends, and reasoning.

     • Conclusion: Actionable recommendations or takeaways."

4. Simplify for Understanding:

   • Use the Reverse Persona Pattern and ask ChatGPT to “explain it in such an easy way that 5th graders easily understand but in a tone fit for a company board meeting.”

   • Focus on straightforward language and relatable analogies without oversimplifying critical details.

   • Template

     "Explain the data and insights in a way that a 5th grader would understand, but maintain a tone suitable for a professional presentation to [specific audience, e.g., board members, stakeholders]. Use analogies and simple language, but ensure all critical details are included."

5. Finalize the Story or Script:

   • Review the script to ensure it’s polished, engaging, and suitable for presentation.

   • Use the script as a foundation for presenting data to diverse audiences.

Lesson 4: Presentations

1. You Can Use:

1. The Visual representations you create (e.g., charts, scatter plots, graphs, etc.).

2. The story or script you created.

3. The data you used.

4. The conclusions you drew or results you reached.

5. The reasoning behind those results and conclusions.

6. The studies and website links that confirm this data.

2. Create an Outline

1. Create your own presentation outline that includes:

   • Number of slides and the detailed information each slide will contain.

   • Template:

     "I am preparing a presentation on [topic]. Help me create an outline for [number] slides. Specify the detailed content and purpose of each slide to ensure it effectively addresses [specific audience, e.g., colleagues, executives, clients]."

2. Or, give ChatGPT the context and the new task, and ask it to create an outline tailored to the audience (e.g., colleagues, executives, or clients).

   • Template:

     "I need a presentation outline tailored for [specific audience, e.g., executives, colleagues, clients] on [topic]. Provide a logical structure, including the number of slides and the detailed content for each slide. The outline should focus on [key objectives, e.g., presenting data, persuading action, teaching a concept] and maintain a professional tone."

3. Choose a Method of Creating

1. Go to Gamma.app or any of the AI presentation creation tools listed in the AI tools guide provided.

2. Either:

   • Insert the full outline and written information into the tool directly.

   • (Preferred):

     1. Prompt ChatGPT to generate detailed information for each part of the outline, one section at a time.

     2. Continue this process for all parts of the outline (e.g., first section, second section, etc.).

     3. Compile the detailed information into a Google Doc, inserting the outline as the first page.

     4. Download the Google Doc as a PDF.

     5. Upload the PDF to Gamma.app and press "Generate."

     6. Manually add the charts and visual representations to their corresponding parts in the presentation.

3. Result: You now have:

   • A complete presentation.

   • A script.

   • Clear and accurate data representation.

Extra Technique

Conversation Training:

1. Use ChatGPT Advanced Voice mode to train and improve your presentation delivery.

Steps:

• Practice presentations by Talking to advanced voice mode.

• Ask ChatGPT to critique your presentation skills based on your Tone, vocabulary etc.

• Use ChatGPT’s feedback to refine your presentation skills.

Lesson 5: Video and Audio Presentations

1. Create a Voiceover Using AI

If you have a story or script created using the storytelling technique, you can easily generate a professional AI voiceover:

1. Go to ElevenLabs or any of the free text-to-voice models listed in the AI tools guide.

2. Paste Your Script into the tool.

3. Choose Your Voice:

   • Select a default voice.

   • Optional: Pay $22 to use your own cloned voice.

4. Generate the Audio by clicking the appropriate button.

5. Download the Audio file once it's ready.

6. Use the audio to:

   • Option 1: Create a video by importing it into a video editing software like CapCut:

     1. Import your presentation slides into the editor.

     2. Sync the audio to the slides by:

        • Listening to the audio.

        • Adjusting the timing of each slide transition to match the narration.

   • Option 2: Play the audio during the presentation and manually transition the slides.

   • Option 3: Send the audio to a team member for communication purposes.

2. Extra Technique: Create a Talking Avatar

You can add a talking avatar to make your presentation more engaging:

1. Use a Voice-to-Video Tool:

   • Go to Synthesia or any text-to-video AI tool listed in the AI tools guide.

2. Insert Audio or Script:

   • Upload your audio file or paste the script directly into the tool.

3. Generate the Video:

   • The tool will create a talking avatar delivering your script.

4. Add the Avatar to Your Presentation:

   • Using CapCut:

     1. Import the video with the avatar.

     2. Use the Mask Circle feature to crop the avatar into a smaller circle.

     3. Place the avatar in the bottom-right corner of your presentation slides.

     4. Export the final video by clicking the Export button.

5. Workflow Example (Global Solutions Inc.)

Continue on your example and create each of the communication techniques

6. Assignment: Communicating Results Exercise

1. Continue on your example

2. Use at least 3 of the communication techniques we recommend you to try all of them.

3. (Optional) Upload them to google drive and share them with us in the comments

7. Key Takeaways & Next Step

• Communicating Results transforms the decisions and data from previous phases into usable instructions, stories, or presentations.

• You can tailor content to each audience (supervisor vs. CFO) and deliver it in multiple formats (emails, slides, voice-overs).

• Tools like Gamma (for presentations) or AI voice generators amplify your reach, while ChatGPT ensures clarity and consistent tone.

Phase 7 — Creative thinking

Now that you can compile data, organize it, generate logical reasoning and make decisions based on Them and communicate them effectively. its time for a little out of the box thinking to add innovation in your job.

Phase 7 – Creative Thinking

In this final phase of the White-Collar AI Workflow, you’ve already gathered data (Phases 1–2), analyzed it (Phases 3), made decisions and confirmed everything’s correct (Phase 4,5), communicated those results effectively (Phase 6). Phase 7 is about pushing beyond the usual solutions and utilizing AIs most powerful feature which is Innovative solutions—brainstorming and coming up with fresh ideas that might give your Firm an innovative edge.

Lesson 1: Unlimited creativity

1. Why Creative Thinking Matters

• Staying Competitive: Even if you solve problems well, new challenges or markets arise. Creativity keeps you one step ahead.

• Continuous Improvement: A bright idea—like a new product spin, a unique marketing angle, or a better shift schedule—can evolve from the data you already have.

• Team Engagement: Employees often feel energized contributing fresh perspectives. AI can spark additional angles, but human creativity remains essential.

Scenario at Global Solutions Inc.

• Marketing wonders if they could combine the short-term ad success with a totally new product approach gleaned from competitor data.

• Operations might dream up a new line configuration or automation triggered by sensor data.

• Finance could see potential expansions in emerging markets, using AI to brainstorm more cost-effective or revenue-boosting ideas.

2. Creative Idea Generation

1. Generate creative ideas:

   • Use ChatGPT to generate creative ideas.

2. Ask for a large number of ideas:

   • Request ChatGPT to generate 100 creative ideas related to a specific topic.

3. Evaluate and select:

   • Review the generated ideas and select the top 3 or the best ones for your needs.

4. Cycle of refinement:

   • using the same concept of the cycle of refinement you can:

     • keep generating ideas

     • have ChatGPT criticize them

     • generate solutions for the criticism points he mentioned till your satisfied with the result

• Template

  "Generate a wide range of creative ideas about [specific topic/problem]. Provide at least [number, e.g., 100] unique ideas, ensuring they cover diverse approaches, perspectives, and potential solutions."

5. Workflow Example (Global Solutions Inc.)

1. Marketing

   • They ask ChatGPT for 15 “beyond normal” ways to cross-promote the new ad campaign with another department.

   • A few are too extreme, but 2–3 are promising. They refine them, then propose a pilot synergy with Finance (like a co-branded email on cost-saving tips tied to product promotions).

2. Operations

   • Brainstorms a “quiet hour” after lunch—where no announcements or meetings happen, letting staff focus. ChatGPT helps weigh pros (more productivity) vs. cons (less real-time communication).

   • They adopt a small test on line #3 to see if daily stress drops, possibly increasing output.

3. Finance

   • Considering expansions into a new region, ChatGPT suggests unconventional cost-sharing deals or partial investments with local partners.

   • Reverse analysis: “What if we avoided expansions and doubled down on domestic sales?” This ensures they’re not missing an even better approach.

Continue on organic fruits example to generate creative ideas for the same workflow

6. Assignment: Creative Thinking Exercise

1. Using the same example you used in the previous 6 phases

   1. Generate out of the box ideas that your company can use based on the data and analysis you conducted with ChatGPT

   2. Choose the best 1 and iterate it till perfection

   3. paste it in the comments below

7. Key Takeaways, What’s Next & Resources

• Creative Thinking ensures you don’t just settle for the solutions found so far. You keep innovating.

Full workflow checklist

PHASE 1: DATA INPUT & GATHERING

• Collect all files and info.

• Attach/Paste text, images, or audio transcripts.

• Check if anything’s missing.

• Ask ChatGPT if it needs more details.

PHASE 2: DATA ORGANIZATION

• Outline or structure the data.

• Spot conflicts or unclear parts.

• Refine any messy info.

• Ask ChatGPT if it needs clarifications.

PHASE 3: DATA ANALYSIS

• State your goal or question.

• Get ChatGPT’s initial analysis.

• Ask ChatGPT to self-critique and refine.

• Keep top solutions or insights.

PHASE 4: CONFIRMING RESULTS

• Double-Check sources and numbers.

• Verify internal vs. external data.

• Review with team if needed.

• Update final facts in ChatGPT.

PHASE 5: DECISION MAKING

• Compare your best options.

• Play out scenarios (short vs. long term).

• Refine until you pick one solution.

• Confirm all angles are covered.

PHASE 6: COMMUNICATING RESULTS

• Summarize the final decision.

• Tailor messages to each audience.

• Use visuals or scripts if needed.

• Share via email, presentations, or voiceovers.

PHASE 7: CREATIVE THINKING

• Brainstorm fresh ideas beyond the basics.

• Evaluate the most promising ones.

• Refine with pros/cons.

• Test a small pilot if possible.

ALWAYS REMEMBER

• Ask ChatGPT for missing details.

• Keep each phase short and clear.

• Repeat self-critique and refinement steps.

• Save final notes or references for quick review.

Final Step: Wrap-Up & Next Steps

With Phase 7 concluded, you’ve seen the entire workflow—from data gathering (Phase 1) through decision-making (Phase 4) and creative expansion (Phase 7).

Your journey is far from over You don’t need to worry we will notify and update you with the newest techniques and lessons when we think it would be worth your attention, for now try and implement what you learned check in every 2 weeks to see the latest action steps and better your career