Transform Problem-Solving with AI using Universal Problem-Analysis Companion©

Discover how a series of AI-powered prompts can revolutionise how you define, analyse, and solve problems.

23 min readDec 31, 2024

Transform Problem-Solving with AI using Universal Problem-Analysis Companion article by Greg Twemlow, image by Ideogram

This article serves as the priority date for the claim of Copyright for the phrase “Universal Problem-Analysis Companion©”, dated December 31st 2024 in Sydney Australia.

Listen to the Deep-Dive Podcast.

Read the Epilogue to see an example of the Universal Problem-Analysis Companion© Prompt process.

Have you ever felt overwhelmed trying to solve a problem, unsure if you’re even addressing the right issue? Tackling problems can be frustrating and time-consuming, whether in the workplace, at home, or in personal projects. We could reflect on the fact that life, in many ways, is just an endless stream of problems to be solved. Given the constant need to solve problems, a reliable, AI-powered framework is a game-changer.

But with the Universal Problem-Analysis Companion©, you’ll navigate these challenges with a sense of control and relief, knowing a structured approach guides you.

This Universal Problem-Analysis Companion© is for everyone — not just for professional analysts or consultants. It’s designed to be accessible, practical, and highly effective. Most importantly, it’s inclusive, welcoming all users and showing how transformative AI can be, even if you’ve never used it. In fact, the Universal Problem-Analysis Companion© is the perfect entree into AI for people who have not yet used AI or have only done rudimentary AI testing.

Why I Created the Universal Problem-Analysis Companion©

In my experience, one of the biggest challenges people face when solving problems is knowing where to start. Too often, we rush into solution mode without clearly defining the problem, which leads to wasted time, effort, and resources. How often have you been on a team that thinks it has solved a problem, only to find the problem returns?

I created this companion to address that challenge head-on. By combining proven problem-solving techniques with AI’s ability to do the heavy lifting, this companion gives you:

Clarity: Focus on defining the real problem, not just the symptoms, while achieving agreement of stakeholder perspectives.
Alignment: Organise diverse perspectives and ensure stakeholder agreement.
Guidance: Use AI to ask the right questions and identify patterns.
Efficiency: Reduce the time needed to analyse, validate, and solve problems.

What the Companion Does

At its core, the Universal Problem-Analysis Companion© is a series of standardised AI prompts that guide users through a structured problem-solving workflow. These prompts act as a dynamic template, enabling a consistent and practical approach to solving any challenge. By running the process entirely within a single ChatGPT session, the companion ensures every aspect of the analysis is captured in one place for seamless continuity.

Why AI Makes All the Difference

The AI-powered nature of this companion is what sets it apart. By embedding standardised prompts within a dynamic, interactive system, the companion transforms how problems are approached and solved:

Fast: The AI handles the heavy lifting, analysing inputs, identifying patterns, and synthesising insights in minutes.
Objective: AI eliminates biases, ensuring a transparent and impartial analysis of each issue.
Scalable: Standardised prompts create consistency, whether you’re solving one problem or many.
Context-Aware: Context is the secret-sauce for success with AI, and the companion gets smarter with each problem, leveraging cumulative insights to improve future analyses.

Who Can Use Universal Problem-Analysis Companion©

This companion is for anyone who wants to:

Solve problems more effectively, whether in their personal life or professional environment.
Experience the power of AI without needing technical expertise.
Work collaboratively with others to achieve clarity and alignment on complex issues.
Prove the transparency of the process and decision logic.

Whether you’re a student, a team leader, or someone who wants to make better decisions, this companion gives you a proven, AI-assisted path to better outcomes.

The First Step to Better Problem-Solving

The Universal Problem-Analysis Companion© is more than just a set of tools — it’s a repeatable approach for anyone needing a “companion or collaborator” to guide their work in solving problems. Combining standardised prompts with AI’s power provides a structured and adaptable framework, transforming how you approach challenges.

Are you ready to unlock the power of AI and take your problem-solving skills to the next level? Let’s get started.

Six Stages of the Universal Problem-Analysis Companion©

Six stages guide users through a structured problem-solving process, leveraging AI to ensure clarity, consistency, and actionable outcomes. Each stage corresponds to the workflow outlined in the Universal Problem-Analysis Companion Prompt Process (see below).

Step 1: Describe the Problem

Purpose: Capture diverse perspectives on the problem.

Step 2: Validate the Root Cause

Purpose: Use structured techniques to identify and agree on the fundamental issue.

Step 3: Assess the Impact

Purpose: Quantify the consequences of the problem.

Step 4: Synthesise Diverse Perspectives

Purpose: Reconcile and integrate additional insights.

Step 5: Propose Starting Points for Solutions

Purpose: Explore potential solutions within defined constraints.

Step 6: Enable Continuous Problem-Solving

Purpose: Identify patterns and systemic opportunities across multiple problems.

How to Use This Template

Run the Process in a Single AI Chat Session

Maintain continuity and retain context throughout the analysis, which can be conducted over hours, days or weeks.

Iterate or Cycle as Needed Before Moving to the Next Prompt

Use follow-up prompts to refine each step and adapt based on stakeholder input.

AI Documents the Results

Your preferred AI records the entire analysis process to enable an independent review of the process and decisions for easy reference and future learning.

The template below provides a robust, AI-driven framework to tackle complex problems systematically and collaboratively. The prompts are adaptable and can be customised based on the nature of the problem or the stakeholder group.

Template to be Copied: Universal Problem-Analysis Companion Prompt Process Instruction

Copy from “Begin Text” to “End Text” below and paste into ChatGPT or your preferred AI to initiate the six-stage Universal Problem-Analysis Companion© process.

Begin Text:

Initiate the Universal Problem-Analysis Companion Prompt Process

Prompt: “Please initiate the six-stage Universal Problem-Analysis Companion Prompt Process.”

This command launches the entire process, guiding the user step-by-step through the six stages of problem analysis. The companion ensures all inputs, clarifications, and iterative refinements are managed within a single, continuous session.

Universal Problem-Analysis Companion Prompt Process

The Six Stages of Problem Analysis

Describe the Problem:

The responsibility here lies with the stakeholders, who often have slightly different definitions of the same problem. The companion starts by capturing each version and labelling it clearly (e.g., “Stakeholder 1,” “Stakeholder 2,” etc.).
Example Prompt: “Describe the problem from your perspective. Be as specific as possible. If there are multiple stakeholders, please include their perspectives separately.”

The AI synthesizes these inputs and recommends a blended problem description incorporating key elements from all perspectives.

2. Validate the Root Cause:

AI supports stakeholders by using techniques like the 5 Whys to drill down to the fundamental issue.
This step often reveals that initial descriptions from Step 1 were incorrect or incomplete. If so, the process circles back to refine the problem definition.
Example Prompt: “Why do you think this is happening? Keep asking ‘why’ until you reach a root cause that explains the issue.”

Decision Point: Stakeholders must agree on the validated root cause before proceeding. Use AI to identify gaps or disagreements and suggest paths to alignment.

3. Assess the Impact:

AI assists in quantifying and ranking the problem’s impact through targeted prompts.
Example Prompts:
“What happens if this problem isn’t solved?”
“Who or what is most affected by this issue?”

4. Synthesize Diverse Perspectives:

If additional perspectives arise during impact assessment, stakeholders revisit the initial analysis with AI guidance to ensure completeness.
Example Prompt: “Do new insights or perspectives alter our understanding of the problem or its impact?”

5. Propose Starting Points for Solutions:

Stakeholders outline constraints that apply to potential solutions, such as budget limits, technology preferences (e.g., cloud-based solutions), or timeline restrictions.
Example Prompt: “What constraints should we consider for potential solutions? For example, is there a budget limit, or must the solution be cloud-based?”

Using these constraints, AI conducts objective research and evaluates solutions it is aware of, proposing recommended options.

This iterative process allows stakeholders to refine constraints and explore additional possibilities with AI’s guidance.

Example Prompts:
“Based on these constraints, outline three potential solutions. For each, explain how it fits within the constraints and its likely effectiveness.”
“Would you like to refine the constraints or explore additional possibilities?”

6. Enable Continuous Problem-Solving:

The companion can handle subsequent problems within the same session for teams or small companies tackling multiple challenges. By retaining context, AI identifies patterns, recurring issues, and systemic opportunities for improvement.
Example Prompt: “Review all previous problems analyzed in this session. Do you see any connections or overarching themes?”

Tips for Effective Use

Session Continuity

If the session is interrupted, you can resume by pasting saved inputs or asking ChatGPT: “Continue the Universal Problem-Analysis Companion process from where we left off.”

Optional Adjustments:

Feel free to adapt the process as needed. For example, if the root cause is already validated, skip to Step 3 by typing: “Move to Step 3: Assess the Impact.”

Clarify Example Prompts:

The companion generates similar prompts automatically. The examples provided are illustrative and may be tailored during the session.

Save Your Results:

Once completed, save or summarize the session for future reference. This ensures learnings and outputs are preserved for later use.

End Text.

Epilogue: here’s an example of the flow of using in ChatGPT

This is a lengthy Chat and I show below a summary of the ten themes of the Chat. I recommend that you read the entire Prompt session to get a sense of how the ChatGPT responses made me think of the practical issues and how my thinking influenced ChatGPT to arrive at a very practical approach that is feasible for the fictitious company I described.

TL;DR — please summarise this Chat into ten key themes of the Chat.
ChatGPT said:
ChatGPT
Ten Key Themes from the Chat
Problem Definition and Opportunity Framing:
The SME’s challenge is to uncover insights from past assignments to enhance consulting impact and client relationships, framed as an opportunity rather than a problem.

Historical Document Analysis Goal:
The focus is analysing documents from assignments since 2015, identifying trends and patterns to improve strategic decision-making.

Initial Constraints and Challenges:
Challenges include inconsistent documentation, lack of a formal repository, and the need to extract insights from varied formats like Word, Excel, PDFs, and emails.

Refined Focus on Recent Documents:
To keep the project manageable, the scope was narrowed to analysing documents from the past 12 months, starting with 50 assignment PDFs.

Systematic Naming and Organization:
A naming convention for PDFs was proposed (e.g., “Client-Project-Date”) to improve organization and facilitate analysis.

AI-Driven Document Analysis Process:
ChatGPT was identified as a key tool for analyzing documents, using specific prompts to extract client objectives, methodologies, outcomes, challenges, and lessons learned.

Pilot Project Implementation:
A pilot program was designed, starting with ten assignments, to test the process and refine it based on insights and team feedback.

Team Involvement and Collaboration:
A staff member was assigned to manage the pilot, ensuring text extraction and analysis were conducted systematically, with team collaboration for reviewing learnings.

Iterative Refinement:
The process includes iterative feedback loops after analyzing the first ten assignments, allowing for adjustments before scaling up to all 50 assignments.

Focus on Future Scalability:
The approach emphasizes scalability, starting small to test tools and processes, with the potential to implement a full-scale centralized knowledge repository later.

This summary captures the core discussions and decisions, ensuring clarity on the project’s objectives and next steps.

Here is my full ChatGPT session:

Initiate the Universal Problem-Analysis Companion Prompt Process

Prompt: “Please initiate the six-stage Universal Problem-Analysis Companion Prompt Process.”

Universal Problem-Analysis Companion Prompt Process
The Six Stages of Problem Analysis
Describe the Problem:
The responsibility here lies with the stakeholders, who often have slightly different definitions of the same problem. The companion starts by capturing each version and labelling it clearly (e.g., “Stakeholder 1,” “Stakeholder 2,” etc.).
Example Prompt: “Describe the problem from your perspective. Be as specific as possible. If there are multiple stakeholders, please include their perspectives separately.”
The AI synthesizes these inputs and recommends a blended problem description incorporating key elements from all perspectives.

2. Validate the Root Cause:

AI supports stakeholders by using techniques like the 5 Whys to drill down to the fundamental issue.
This step often reveals that initial descriptions from Step 1 were incorrect or incomplete. If so, the process circles back to refine the problem definition.
Example Prompt: “Why do you think this is happening? Keep asking ‘why’ until you reach a root cause that explains the issue.”
Decision Point: Stakeholders must agree on the validated root cause before proceeding. Use AI to identify gaps or disagreements and suggest paths to alignment.

3. Assess the Impact:

AI assists in quantifying and ranking the problem’s impact through targeted prompts.
Example Prompts:
“What happens if this problem isn’t solved?”
“Who or what is most affected by this issue?”
4. Synthesize Diverse Perspectives:

If additional perspectives arise during impact assessment, stakeholders revisit the initial analysis with AI guidance to ensure completeness.
Example Prompt: “Do new insights or perspectives alter our understanding of the problem or its impact?”
5. Propose Starting Points for Solutions:

Stakeholders outline constraints that apply to potential solutions, such as budget limits, technology preferences (e.g., cloud-based solutions), or timeline restrictions.
Example Prompt: “What constraints should we consider for potential solutions? For example, is there a budget limit, or must the solution be cloud-based?”
Using these constraints, AI conducts objective research and evaluates solutions it is aware of, proposing recommended options.

This iterative process allows stakeholders to refine constraints and explore additional possibilities with AI’s guidance.

Example Prompts:
“Based on these constraints, outline three potential solutions. For each, explain how it fits within the constraints and its likely effectiveness.”
“Would you like to refine the constraints or explore additional possibilities?”
6. Enable Continuous Problem-Solving:

The companion can handle subsequent problems within the same session for teams or small companies tackling multiple challenges. By retaining context, AI identifies patterns, recurring issues, and systemic opportunities for improvement.
Example Prompt: “Review all previous problems analyzed in this session. Do you see any connections or overarching themes?”
Tips for Effective Use
Session Continuity

If the session is interrupted, you can resume by pasting saved inputs or asking ChatGPT: “Continue the Universal Problem-Analysis Companion process from where we left off.”

Optional Adjustments:

Feel free to adapt the process as needed. For example, if the root cause is already validated, skip to Step 3 by typing: “Move to Step 3: Assess the Impact.”

Clarify Example Prompts:

The companion generates similar prompts automatically. The examples provided are illustrative and may be tailored during the session.

Save Your Results:

Once completed, save or summarize the session for future reference. This ensures learnings and outputs are preserved for later use.
ChatGPT said:
ChatGPT
Universal Problem-Analysis Companion Prompt Process
Stage 1: Describe the Problem

Please provide the problem description from your perspective. If there are multiple stakeholders, share each perspective separately, and I will synthesize these inputs into a blended problem description. Be as specific as possible.

Example Prompt:
“Describe the problem from your perspective. If multiple stakeholders are involved, include their perspectives separately (e.g., Stakeholder 1, Stakeholder 2, etc.).”

Let’s start with this. Could you describe the problem and any relevant context?

You said:
The context is my SME employing ten staff. We are a human resource consulting business. I need to find a way to reveal learnings from all of our assignments since the business started in 2015. I am hoping that analysing all the documents related to every assignments we will discover trends and insights.
ChatGPT said:
ChatGPT
Universal Problem-Analysis Companion Prompt Process
Stage 1: Describe the Problem
The problem centers around effectively analyzing your SME’s historical assignment documents (2015–present) to uncover trends, patterns, and actionable insights. Each staff member may have different views on the key challenges or benefits of undertaking this task.

Stakeholder Perspectives (Example Input)
Stakeholder 1 (Business Owner):
“I need to unlock learnings from past assignments to identify trends and improve decision-making. However, I’m unsure how to systematically analyze and extract meaningful insights from our existing data.”
Stakeholder 2 (Staff Member 1):
“We store documents inconsistently, making it hard to know where to start. It feels like a monumental task to organize everything.”
Stakeholder 3 (Staff Member 2):
“We have learned a lot from assignments, but those lessons are scattered in individual reports. A central repository or thematic analysis could help us focus on areas like client preferences or successful strategies.”
Synthesis of Problem Statement:
The SME faces challenges in systematically analyzing its historical assignment documents from 2015 onwards to extract trends, patterns, and insights. Key hurdles include inconsistent document organization, lack of a structured analysis framework, and limited visibility into how to connect past learnings with future strategy.

Does this synthesized problem statement align with your view? If adjustments or additions are needed, let me know, and I’ll refine it.

Stage 2: Validate the Root Cause
To proceed, let’s identify the root cause of this challenge. Answer the following questions:

Why do you think the insights from past assignments have not been uncovered yet?
What are the key barriers — e.g., inconsistent documentation, lack of analytical tools, or staff time constraints?
Why is it critical to solve this now?
Once we validate the root cause, we can move to assess the impact.

You said:
I think the correct problem definition is a blend of all three stakeholder perspectives.
ChatGPT said:
ChatGPT
Refined Problem Definition
The SME seeks to systematically analyze historical assignment documents from 2015 onwards to uncover trends, patterns, and actionable insights that can inform strategy and improve decision-making. Challenges include inconsistent document organization, a lack of centralized access to knowledge, and an absence of a structured framework or tools for thematic analysis. This problem impacts the SME’s ability to leverage past learnings effectively and align future actions with proven strategies.

Stage 2: Validate the Root Cause
Let’s identify the root cause of this problem. Answer the following questions to refine our understanding:

Why do you think past insights have remained hidden or unused?
(E.g., inconsistent data storage, lack of time for reflective analysis, or reliance on informal knowledge-sharing.)

What systems or processes currently exist for documenting and analyzing assignments?
(E.g., are documents saved digitally, is there a standard format, or are tools like Excel or project management software in use?)

Why has this problem become critical now?
(E.g., market competition, need for differentiation, or scaling challenges.)

If we uncover gaps in understanding or agreement on the root cause, we’ll loop back to refine this stage. Once validated, we’ll assess the impact of the problem.

You said:
I agree with the examples of point 1 above. We generally conduct the consulting work and close off the assignment to move onto the next assignment.

We don’t have a formal process of collecting and saving all the artifacts related to an assignment. We use all the standard documents like MS Word, Excel, PDFs, and a lot of emails.

I feel it’s not so much a problem as an opportunity that AI could help us realise substantial benefits.
ChatGPT said:
ChatGPT
Updated Problem Definition with Opportunity Framing
The SME has an untapped opportunity to systematically analyze historical assignment documents and artifacts from 2015 onwards to uncover trends, patterns, and actionable insights. These insights could enhance decision-making, refine strategies, and identify new opportunities for growth. Current challenges include inconsistent documentation practices, lack of a formal repository for assignment artifacts, and the manual effort required to extract insights from a mix of formats such as MS Word, Excel, PDFs, and emails. Leveraging AI could enable a structured, scalable approach to realizing these benefits.

Stage 3: Assess the Impact
Let’s explore the potential impact of seizing this opportunity and addressing the underlying challenges. Answer the following:

What benefits would result from uncovering trends and insights?
(E.g., better client engagement strategies, identification of successful consulting frameworks, or more targeted service offerings.)

What happens if this opportunity isn’t pursued?
(E.g., missed growth opportunities, continued reliance on ad-hoc knowledge sharing, or inefficiencies in replicating successful strategies.)

Who or what is most affected by this issue or opportunity?
(E.g., decision-makers, frontline staff, or clients benefiting from more informed strategies.)

With this information, we’ll synthesize diverse perspectives (Stage 4) and move to propose starting points for solutions (Stage 5). Let me know your thoughts!

You said:
re 1 — I think there is likely many learnings from past work that can be applied to enable us to inc rease the impact of our consulting work and help forge stronger connections with our clients.

re 2 — I feel it would be a missed opportunity to keep operating as we have since 2015. AI gives us a Kairos moment that I am very eager to exploit.

re 3 — most affected are me and all my staff because I know that our collective knowledge will be immensely valuable to my company.
ChatGPT said:
ChatGPT
Stage 3: Synthesized Impact Assessment
Benefits of Uncovering Trends and Insights:
The SME is likely to discover numerous valuable lessons from its past assignments. These could directly enhance the impact of consulting work, enable the team to replicate successful strategies, and forge stronger connections with clients by tailoring services to their needs more effectively. These insights could also position the SME as a more data-driven, forward-thinking consultancy in the market.

Consequences of Not Pursuing This Opportunity:
Continuing operations as before would mean missing a unique, time-sensitive opportunity to leverage AI as a transformative tool. This Kairos moment — seizing the right opportunity at the right time — could allow the SME to achieve significant operational and strategic improvements. Missing this would risk stagnation and perpetuate inefficiencies in knowledge application.

Who is Most Affected:
The business owner and all staff are directly impacted, as the collective knowledge embedded in past assignments represents a goldmine of value. By harnessing this knowledge, the entire team could work more efficiently and strategically, achieving greater impact and satisfaction in their roles.

Stage 4: Synthesize Diverse Perspectives
Based on your inputs, let’s ensure we’re capturing a holistic understanding:

Internal Knowledge Value: The untapped collective insights from assignments can transform consulting practices and enhance client relationships.
Strategic Opportunity: AI provides a timely, powerful means to uncover and systematize these insights.
Team Empowerment: Both the business owner and staff stand to benefit from better access to shared learnings, driving professional growth and company success.
Stage 5: Propose Starting Points for Solutions
To move forward, we need to identify constraints and brainstorm possible solutions. Here are some steps and prompts to guide the process:

Constraints and Preferences:

Are there budget limitations for investing in AI tools or consultants?
Do you prefer a specific platform or technology (e.g., cloud-based solutions)?
What is your timeline for implementing this solution?
Potential Solutions:

AI Document Analysis: Use an AI tool like OpenAI, Azure Cognitive Search, or similar solutions to extract insights from your documents automatically.
Centralized Knowledge Repository: Create a structured repository where all future and past assignment data can be stored and tagged for easy retrieval.
AI-Powered Summarization: Use AI tools to summarize and synthesize learnings from assignment documents.
Thematic Analysis Framework: Develop a taxonomy of themes or trends (e.g., client needs, success factors) and train an AI model to identify these in your documents.
Would you like to explore one of these solutions or refine the constraints?

You said:
re 1 — I don’t have a specific budget however I’m prepared to invest funds to achieve the objectives. I could allocate up to $50,000 over a 6-month period. We aren’t a high-tech company so I need advice on suitable technology. The timeline ideally is the next 6-months.

re 2 — I feel like the Pareto Principle applies in that we can likely find 80% of all the files from past assignments. The biggest challenge is how to capture all the knowledge from emails related to our work. I was hoping the Centralized Knowledge Repository would be AI or perhaps a private GPT. I assume we;ll use our ChatGPT account “to summarize and synthesize learnings”. Likewise I am hoping that ChatGPT will uncover “a taxonomy of themes or trends”.
ChatGPT said:
ChatGPT
Stage 5: Refined Constraints and Starting Points for Solutions
Refined Constraints:
Budget: Up to $50,000 allocated over six months.
Timeline: Solution needs to be implemented and generating results within six months.
Technology Preferences: SME-friendly tools that don’t require extensive technical expertise, preferably cloud-based solutions or private GPT instances.
Primary Objectives:
Identify 80% of files from past assignments.
Effectively capture knowledge from related emails.
Centralize this knowledge into a repository for easy access and analysis.
Leverage ChatGPT for summarization, thematic analysis, and insight discovery.
Potential Solution Pathways
**Centralized Knowledge Repository with AI Integration
Steps:

Use a document management tool like Notion, Airtable, or Google Workspace as the repository.
Integrate AI-powered tagging and search using tools like Azure Cognitive Search, Elastic Enterprise Search, or Docugami.
Automate ingestion of assignment documents and email threads into the repository.
Benefits:

Provides a single source of truth for all assignment artifacts.
Enables quick retrieval and thematic categorization of documents.
Ensures all documents are stored in a structured and searchable format.
Email Extraction and Analysis
Steps:

Use tools like Mailparser or Zapier to extract relevant emails and save them as text or PDF files.
Utilize ChatGPT to process and summarize email threads, extracting key insights and decisions.
Challenges:

Requires filtering to capture only assignment-related emails.
Some manual effort may be needed to clean up the data.
Private GPT Instance for Analysis
Steps:

Deploy a private GPT model using OpenAI’s API, or tools like Anthropic Claude or LlamaIndex.
Train it on your historical documents and emails to perform tailored analysis.
Use it to generate a taxonomy of trends, key client preferences, and successful frameworks.
Benefits:

Customizable to your specific needs.
Ensures data security and privacy.
Thematic Analysis with ChatGPT
Steps:

Use your ChatGPT account to analyze documents and extract recurring themes.
Develop a taxonomy collaboratively, starting with ChatGPT suggestions and refining with staff input.
Challenges:

May require iterative refinement to ensure meaningful categorization.
Combination Approach (Recommended):
Centralized Repository: Start by creating a repository (Airtable/Google Workspace).
Automated Email Ingestion: Use Mailparser or Zapier to capture relevant emails.
AI-Powered Analysis: Train a private GPT instance or use ChatGPT API to analyse the repository.
Taxonomy Refinement: Collaborate with ChatGPT and staff to develop actionable themes and insights.
Next Steps:
Choose a Repository Platform: Airtable, Notion, or Google Workspace.
Evaluate Email Ingestion Tools: Identify the best fit for automating email capture.
Explore Private GPT Deployment: Research providers like OpenAI API, Anthropic, or Azure OpenAI Services.
Draft a 6-Month Implementation Plan: Prioritize repository setup, email integration, and AI analysis.
Would you like assistance with drafting the implementation plan or evaluating specific tools?

You said:
That sounds complex and expensive. Could we devise a test program that looks at documents from the past 12-months to get ideas on what trends could be extracted? Just concentrate on documents for now. WDYT?
ChatGPT said:
ChatGPT
Focusing on a test program for the past 12 months is a smart way to keep things manageable, cost-effective, and exploratory. By narrowing the scope, you can evaluate potential insights and refine the process before scaling up. Here’s a streamlined approach tailored to your goals:

Test Program Plan: Extracting Trends from the Past 12 Months
Objective:
Analyze assignment-related documents from the past 12 months to identify trends, patterns, and actionable insights using AI-assisted tools in a controlled, low-cost trial.

Step 1: Define the Scope
Document Types: Focus on key artifacts such as:

Client proposals and reports (MS Word, PDFs).
Data analyses (Excel files).
Meeting summaries or debrief notes.
Volume: Limit to a realistic subset, e.g., 10–20 representative assignments.

Step 2: Set Up Tools and Frameworks
Centralized Storage for Documents:

Use a free or low-cost tool such as Google Drive, Notion, or Airtable.
Organize documents by client, project, and date.
AI Document Summarization and Thematic Analysis:

Use ChatGPT (your existing account) or free-tier tools like OpenAI Playground.
Upload document text or paste key excerpts into ChatGPT for summarization.
Ask AI to identify recurring themes, e.g., client needs, successful strategies, challenges faced.
Example Prompt for ChatGPT:
“Analyze this document and summarize the main themes or patterns related to client objectives and outcomes.”

Low-Cost Text Analysis Tools (Optional):

Tools like MonkeyLearn or Lexalytics for identifying trends in text data.
Step 3: Conduct Analysis
AI-Assisted Insights:

Process each document or a batch of documents through ChatGPT.
Collect AI-generated summaries and group recurring themes.
Manual Review (if needed):

Have staff verify and refine AI insights to ensure relevance.
Step 4: Identify Trends and Patterns
Create a simple taxonomy based on themes identified, e.g.,:

Common client challenges.
Repeatedly successful approaches.
Emerging client preferences or opportunities.
Present these insights to the team for feedback and refinement.

Step 5: Evaluate the Pilot’s Success
Measure success based on:

Number of useful trends or patterns identified.
Applicability of insights to future assignments.
Feedback from staff on ease of the process.
Use this evaluation to decide whether to:

Expand the analysis to more documents.
Invest in more advanced tools for automation.
Build a full-scale knowledge repository.
Benefits of This Approach
Low Cost: Leverages existing tools (e.g., ChatGPT, Google Drive).
Quick Insights: Focuses on a small, manageable dataset.
Scalability: Provides a framework to expand the process if it proves valuable.
Practical Outcomes: Enables immediate application of identified trends in current projects.
Would you like assistance drafting specific prompts for analysing the documents or organising the workflow for this pilot?

You said:
I think that we can create a single PDF document of the assignments from the past year, and we can name the PDF using “client, project, and date”, like “Acme Buttons-Staff Upskilling-Mar 2023”. I’ll limit the PDFs to 50 of the assignments from the past year. I need your help to create the Prompt for ChatGPT to analyze the 50 PDFs. WDYT?
ChatGPT said:
ChatGPT
Creating a single PDF for each assignment and naming it systematically is an excellent strategy for organization. Analyzing 50 PDFs using ChatGPT can uncover trends and insights effectively. Below is a proposed prompt structure and workflow for analyzing each PDF:

Prompt Structure for Analyzing Each PDF
Base Prompt for Single PDF Analysis:
_”Analyse the content of this document. Summarize the following key elements:

The client’s main objective or problem statement.
The approach or methodology applied to address the objective.
Key outcomes or results achieved.
Challenges or obstacles encountered during the assignment.
Any lessons learned or opportunities for future improvement.”_
Optional Addition:
“If applicable, identify recurring themes or trends related to the client’s industry, needs, or challenges.”

Batch Summary Prompt (After All PDFs Are Processed)
Once the individual PDFs are analysed, you can ask ChatGPT to identify overarching trends and insights from the collective summaries:

Batch Analysis Prompt:
_”Based on the summaries of 50 assignments, identify recurring themes, patterns, and insights, including:

The most common client objectives or problems.
Frequently applied approaches or methodologies.
Repeatedly successful outcomes or strategies.
Common challenges faced across assignments.
Emerging trends or opportunities that could guide future consulting work.”_
Workflow for Processing PDFs
Prepare PDFs:

Organize the 50 assignment PDFs with clear naming conventions (e.g., “Acme Buttons-Staff Upskilling-Mar 2023”).
Extract Text from PDFs:

Use a PDF-to-text extraction tool like Adobe Acrobat, Smallpdf, or PDF Extractor to convert each PDF into readable text for ChatGPT.
Analyze Each PDF in ChatGPT:

Input the extracted text into ChatGPT using the single-PDF prompt.
Save the output summary for each assignment (e.g., in a shared Google Doc or spreadsheet).
Conduct Batch Analysis:

Compile all individual summaries into one document.
Use the batch analysis prompt to identify overarching insights.
Tips for Efficient Analysis
Break It Into Phases: Process and analyze 10 PDFs at a time to avoid being overwhelmed and allow iterative refinement of prompts.
Focus on Key Sections: If a document is lengthy, extract and analyze only the sections most relevant to the analysis (e.g., executive summary, results).
Save Outputs Systematically: Create a folder for ChatGPT-generated summaries, naming each summary file after the corresponding PDF.
Next Steps
Would you like me to refine the prompts further, or assist with suggestions on PDF-to-text tools and methods to streamline this process?

You said:
I will allocate one of my staff to manage the process you have suggested. She will start as you suggest with PDFs of ten client engagements. I think she will be able to do text extraction. Since we have a team ChatGPT account the whole team will be able to follow the process. Once we have processed ten assignments we’ll meet as team to discuss the process and learnings.

Thanks for your help.
ChatGPT said:
ChatGPT
That sounds like a practical and well-structured plan! Allocating a staff member to manage the process and starting with a smaller batch of ten client engagements is a smart way to test the approach before scaling up. Here’s a quick recap to ensure everything is streamlined:

Steps for Your Staff Member:
Prepare the First Ten PDFs:

Organize and name the files using the agreed convention (e.g., “Client-Project-Date”).
Ensure text extraction is complete and clear.
Run Individual Analyses in ChatGPT:

Use the provided single-PDF prompt to analyze each document.
Save the output summaries systematically, such as in a Google Doc or spreadsheet.
Compile Summaries:

Combine the ten summaries into one document for ease of review.
Highlight Key Insights:

Use the batch analysis prompt to identify trends or themes from the ten assignments.
Team Meeting Agenda:
Review the Process:

Discuss what worked well and any challenges in managing the PDFs or running the analyses.
Evaluate Insights:

Examine the trends and themes identified from the first ten assignments.
Discuss whether the insights align with team expectations or suggest new directions.
Plan for Scaling Up:

Decide whether to adjust the process before analyzing the remaining 40 assignments.
Discuss opportunities to improve organization, prompts, or knowledge-sharing strategies.
Your team is on track to uncover valuable insights while iteratively refining the approach. If you need further assistance during or after the pilot phase, feel free to ask. Best of luck with your project — I’m confident it will lead to meaningful improvements in your consulting work!