Q 793. In many AI-enabled workflows, the prompt is the actual instruction that guides the model’s behavior — from extracting insights to responding to customers. A poorly framed prompt can lead to confusion, hallucination, or biased responses, even when the model itself is powerful. Think of a process in your domain where AI is expected to make or support a decision. How could the quality of prompt design influence the accuracy, tone, or trustworthiness of the outcome? Share an example (real or hypothetical) where a small change in prompt wording could make a big difference.

 

The best answer will be selected on the basis of: 

• Clarity of the process and decision involved
• Insight into how prompt design affects results
• Practicality and creativity of the suggested prompt improvement

 

Note for website visitors -

• This platform hosts two weekly questions, one on Monday and the other on Thursday. 
• All previous questions can be found here: https://www.benchmarksixsigma.com/forum/lean-six-sigma-business-excellence-questions/. 
• To participate in the current question, please visit the forum homepage at https://www.benchmarksixsigma.com/forum/.
• The question will be open until Monday or Thursday at 5 PM Indian Standard Time, depending on the launch day. 
• Responses will not be visible until they are reviewed, and only non-plagiarised answers with less than 5-10% plagiarism will be considered for winner selection. 
• If you are unsure about plagiarism, please check your answer using a plagiarism checker tool such as https://smallseotools.com/plagiarism-checker/ before submitting. 
• All correct answers shall be published, and the top-rated answer will be displayed first. The author will receive an honourable mention in our Business Excellence dictionary at https://www.benchmarksixsigma.com/forum/business-excellence-dictionary-glossary/ along with the related term.

• Some people seem to be using AI platforms to find forum answers. This is a risky approach as AI responses are error-prone because our questions are application-oriented (they are never straightforward). Have a look at this funny example - https://www.benchmarksixsigma.com/forum/topic/39458-using-ai-to-respond-to-forum-questions/

• We also use an AI content detector at https://quillbot.com/ai-content-detector. Only answers with less than 45-50% AI-generated content will be considered for winner selection.

Proper prompt Engineering to LLMs is equivalent to having someone speaking properly to me in my native language.  The better someone can speak my native language and with proper inflections in their enunciations and tone, the better I will understand exactly what they mean to say.  If they can't, it leaves me more up to trying to interpret what they are trying to say.  We've all been in that situation where we weren't clear what someone was trying say and we completely get them wrong.  It's not necessarily our fault its just that we aren't getting clear communication.  No difference when we can't speak (prompts) the native way LLMs can understand.  If we speak poorly, the LLM has to more interpreting. So, It is up to us if we want the LLM to completely understand what we are asking for.  If we don't, we have to expect that the LLM will easily misinterpret, be confused, think that we may be bias one way or another, and making it unclear for the LLM to answer with a precise response. 

 

So yes, expect the output of an LLM to never be reliable, concise or predictable if we don't know how to speak (prompt) in such a way where there is no ambiguity between what I am asking for and what the LLM understands.  It's not difficult for us to imagine any use case where AI is used that things can go horribly wrong very easily and quickly.     

During the CAISA training from Benchmark , I got  introduced with the term "PROMPT" which is a basic requirement during the designing of the AI solution.  The instructions included in the prompt defines the workflow and cycle of the AI solution. Incomplete steps or any Gap in the prompt may lead to non-functional model or inconsistent results to the queries. 

During the practical hands on working on an AI solution,  it was evident that on changing the prompt instructions the results/outcome were varied and also the required outcomes were easily achievable. Prompt design is the heart or the backbone of any AI solution

Using myself as a practical example pre my knowledge of prompt engineering and my use of ChatGPT, I have used AI to write content scripts, respond to emails, draft the flow of strategy sessions for a business leader's workshop, design SOPs, training modules, and presentation scripts in a global manufacturing context. I can tell for a fact that the results I have gotten post my knowledge of prompt engineering and design shows how prompt design can elevate or limit the outcome.

 

My initial prompts was something like "Create an SOP for the Logistics team" or "What caused Line 2's downtime yesterday" but has now evolved into "on yesterday’s downtime on Line 2, using available sensor logs, operator notes, and maintenance history, identify the most probable root causes and suggest low-cost, high-impact fixes relevant to skincare batch making process" or "Create a logistics SOP for a skincare manufacturing plant in Nigeria, including job grade responsibilities, escalation paths, tool access levels, and cross-functional dependencies.” 

 

I think the prompts given to an AI model is critical because it determines whether the output is surface-level or genuinely actionable and within the context of the problem you are trying to solve. The improvement didn’t just help me get better results; it helped the AI model understand my context more deeply.

 

In other words, better prompts led to smarter AI support which in turn leads to better decision making.

Prompt Design 100% will impact AI enabled take decision

Therefore, The quality of prompt design will significantly influence the accuracy, tone, and trustworthiness of AI decision.

In our manufacturing plant, we use prompt and by this prompt we enable AI take a decision to accept or reject any customer complaint received.

So, the quality of prompt is very important in AI decision

A proper designed prompt will guide the AI model to produce relevant, accurate, and unbiased responses, while a poorly designed prompt will lead to confusion, wrong, and biased responses.

AI model is used to analyze any customer feedback or complaint based on prompt designed.
 

Poor of designed Prompt can lead to one or more of the following:

1)      Wrong analysis of the customer complaint or feedback

2)      Reject the correct claim, this will lead customer unsatisfied

3)      Accept the wrong claim, cost the organization

4)      AI model might incorrectly classify the customer complaint/feedback

 
A small changes in prompt can make a big difference like following:
 

1)      To be specific

Adding specific keywords to the prompt will help the AI model focus on specific task.
2) Tone language.
3) Use historical data and how to guide the AI model

To develop good design of prompt, we can use the following guidelines:
 

-          Clearly define the task

-          Provide relevant context to help the AI model understand the task and produce accurate responses.

-          Use specific language

-          Test and refine the prompt design to ensure accurate response.
 

Designing prompts is an important and critical step in getting most out of AI large language-based tools.

For example, healthcare BPO  claims process, let’s consider a scenario where we want to LLM model to review the claim which was denied. We are expecting the LLM model/ AI agent to detail the reason for claim denial, suggesting the next appropriate action and AI agent to execute the action (route the claim to concerned team if requires)

Initial Prompt design: ““Explain this claim denial and next action to be performed”.

 AI model may give a generic explanation like:
“Claim is denied with CO-22 denial code meaning the claim is not covered by the current payer. You should escalate it.”

While this response is correct, it is not clear exact action that agent has to take which may lead to misrouting or unnecessary escalation, and this does not inspire trust in the LLM reliability

Improved Prompt Design:

“The claim has been denied with code CO-109. Explain in 1-2 sentences what this denial means, identify if this is related to member eligibility or provider mismatch, and suggest the most appropriate internal team to route the claim to basis historical resolution logic.”

LLM output basis improved prompt:
The AI gives a more accurate, trustworthy, and actionable response:
“Denial code CO-22 indicates another health insurance payer may be primarily responsible for covering the service. The current payer believes they are not the primary insurance, and payment responsibility may lie with another insurer. Based on historical data, this should be routed to the Coordination of Benefits (COB) / Member Eligibility Verification team to verify COB. Next actions to be performed are:

a)      Review member’s COB file in the system.

b)     Check for existence of other insurance coverage (via client portals, IVR, or eligibility tools).

c)      Update COB order if incorrect.

d)     Coordinate with member or client for COB attestation if required.

e)     Route the claim back to processing once correct COB details are confirmed

 

Now this is a detailed LLM output which AI agent can action with precision and high accuracy. This way, even in a highly regulated environment such as healthcare BPO, a small change in the prompt wording can transform a AI output from ambiguous to precise

In the Denial Management process in RCM (Revenue Cycle Management), we have been trying out few AI tools to help in understanding payer denial reasons and to suggest the team what to do next. The idea is to improve the speed of the resolution, but we realized something important: how can we ask the AI makes a big difference.

 

For instance, if we just say something like “Tell us what this denial means”. The response we get is usually pretty vague, maybe it will say, “Coverage issue” and that’s about it. Not really useful.

But then we tried changing the prompt slightly to “What should a denial analyst do if the remark says Coverage terminated prior to date of service?”

 

That change made the AI give a much better answer. It said to verify the patient’s eligibility, and if the coverage really was inactive on that date, then update the insurance or transfer the balance to self-pay. That’s the kind of reply our teams can actually use.

 

It sounds small, just a change in wording, but it totally changed the quality of the result. We went from something high and border to something actionable.

 

So we design the prompt really affects how accurate and helpful the AI is. It is not about what the process required, it is about how we ask it. That is something we will definitely keep in mind as we use more of these tools.

Prompt plays the important role in quality & Accuracy of AI decision making.

 Context enhances understanding
o AI uses prior information in the prompt to generate context aware responses.
 Clarity leads to accuracy
o Ambiguous prompt produces vague result which can be misleading
 Structure encourage logical flow
o Structure prompt lead to step to step solution
 Specificity Reduces Hallucination
o Vague prompts can mislead the AI .
The results show that well-designed, context-rich prompts significantly improved decision quality, while vague or overly technical prompts led to misinterpretation and delayed responses
 

Prompt design basically defines the character of the AI system. If we take movie making as an analogy, Prompt desiging is like a director of a movie who defines the characters of the actors as per the story. Without proper definition of a character the role of that actor would be flat and look out of place .The character will not display appropriate emotions, dialogue delivery, subtle changes in body language etc., for different scenes, ultimately resulting in the failure of the movie due to a big disconnect between the audience and the characters.

Hence the quality of a prompt design will directly impact the quality of the output of AI.

 

When we consider a conversational BOT, though it can understand Natural language and process it, how it responds as per the defined role will play a major role in the AI's usability and experience that user gains. 

 

Let's say for example a 10 year old is eager to know the Theory of relativity. If the AI responds too technically that not even a single word makes sense to that 10 year old, then in the eyes of the 10 year old AI is not of any help to him. Instead, if the prompt design had included to check the age of the requestor or use appropriate analogy to explain concepts as per age of the requestor, then it would have been more useful.

 

Another example, say a shop floor manufacturing supervisor who enters the production data into the ERP systems, finds few keys on his key pad non functional. So he chats with company's IT support BOT and types " some keys not working in the keyboard", the BOT responds by asking what is the OS version, what is the system Tag, when was the last software patch done etc..The user just shuts the BOT, picks up the phone and talk directly to IT support personnel..The very purpose of the IT support chat BOT failed. 

Here, the BOT should know that it is interacting with a non technical employee and has to converse in his language, ask information which he is capable of giving. There is a famous quote in Marketing, 'always talk the language of the customer ". 

 

When computers were boom in the business world, it started with personal computer, slowly it evolved to personalized computing. The same way AI is just entering to business world, there is no perfect system, but there must be a robust process put in place to monitor, gather feedback and improve the systems, in case of conversational BOTs , improve the prompt. 

 

Again back to movie analogy, the result of a movie be it success or failure depends on how well the audience recieve the product. Successful or not, by having a feedback mechanism it helps the director to learn and make improvements in his next movie. Again it is a cycle of iterations that is applicable in every aspects of life. The same is applicable to AI system as well. Like how a director cannot ask an aged character to dance difficult steps in a movie, the prompts too must be specific, formatted, be on the context and talk the language of the user. 

 

In my view, the generalised AI will slowly be evolved into personalized AI in the coming years. The AI will develop capabilities to understand about the different users and serve them as per their personalized requirement.

 

In summary, The prompt design (continuously evolving and improving) will make the AI descisions more relevant, useful and will serve the purpose for which it was designed.

Problem Statement:

 

Extracting more than 400 invoices from group mailboxes is manual task and time consuming.

 

I am planning to introduce this use case in my organization using AI solution. This more popular AI solution introduced by many organizations. With the help of no code or low code AI available solutions, invoices can be extracted from group mailboxes and saved on OneDrive.  It can also help AP team to extract invoice details like vendor name, invoice number, date, amount and save this data on excel spreadsheet for review on daily basis.

 

Quality of the prompt will have direct impact on the tool operation and quality of the data. If they are not properly used then it will not will not derive expected business results and dissatisfaction in end users and management.  

 

Vague prompt:

 

“Extract invoices form the shared Mailbox [[email protected]] and save them in in OneDrive [Path] on daily basis at 8:00 AM. Collect invoice details and update spreadsheet saved on One Drive [Path] ”

 

Improved Prompt:

 

At 8:00 AM daily, scan the shared mailbox [[email protected]] for emails received after 8:00 AM on the previous day.

Filter Criteria:

The email subject line must contain the keywords: “invoice” or “bill”, OR The email must include a PDF attachment that appears to be an invoice.

Actions:

If the email does not meet the above criteria, mark it as unread.

If the email qualifies, move it to the [AP Invoices] folder within the group mailbox.

Extract the PDF attachment and save it to OneDrive at the following path: [Path].

From each PDF, extract the following details:

Vendor Name

Invoice Number

Invoice Date

Invoice Amount

Save the extracted data into an Excel spreadsheet located at: [Path].

Notification:

Once the task is completed, send a summary email to [[email protected]] with a list of vendor invoices saved to OneDrive.

While a well designed prompt helps to improve the accuracy, tone or trustworthiness of the outcome of the workflow, a poorly designed prompt may result in inaccurate outputs or biased replies even if the underlying model is very powerful.
Use Case - Build an AI assistant that can review the logs and summarize potential risks in a data warehouse migration program between Oracle and Databricks.
A poorly framed prompt like " Consolidate the migration process" will generate outcomes like - listing out the various processes like schema migration, code conversion, data migration , validation etc. But the outcome will not help us to summarize the potential risks. It will ignore the errors, will not capture any actionable insights and will fail to support decision making process.
But upon improving the prompt and calling out explicitly - "List the potential risks with regards to schema mismatches,data integrity issues, performance issues " the AI assistant will scan through the logs and provide relevant details for schema mismatch like - loss of precision between number datatypes; missing timestamp value in Databricks 'date' column vs 'date' in Oracle. These details help to improve the accuracy, trustworthiness of the AI assistant, thus helping to deliver better results.

This response is a summary that fits the requirements:

The ESG Risk Assessment of the Investment Analysis Process and Decision Making indicates that more and more organizations in the financial services industry are utilizing AI models to help them with challenges that have to do with the environment, society, and governance. They gather and organize a lot of unstructured data from other people, like news stories, company reports, and ESG evaluations. These AI-generated summaries could assist analysts figure out how to score companies on crucial ESG parameters and what risks they might face. They can also explain their clients where the data came from, how it wasBN was collected, and what it was used for.

People often write down the ESG problems that a company has had to deal with in the past year. Next, consider about how these kinds of worries could make investing more dangerous.

The way a prompt was made affects its reliability, tone, and accuracy.
If you don't make a prompt appropriately, it could lead to:

You are making a general statement when you say "Company X has environmental problems" without stating what or when.

People say that a model is hallucinating when it thinks there are problems that aren't in the data.

For instance, if the analyst hears too many accusations, it might influence how they think.

You can't tell Company X to "write down all the ESG problems they've had to deal with lately."

Things that are wrong:

The term "all" makes it sound like everything is there, which the model needs to know.

What do you mean when you say "problems"? Are you talking about threats to your reputation, claims, or problems that have been fixed?

"From August 2024 to July 2025, write a summary of verifiable ESG controversies about Company X that were reported in credible news sources or regulatory filings." This could be a better suggestion. Please tell us what the claims are, what happened, and how the company responded. Use language that isn't biased to figure out what could go wrong.

Why It's Better:

If you write down what sort of source it is and when it happened, it will be easier to grasp.

It asks for proof at different levels, which makes it more reliable.

Shows how to change the tone of a summary to make it sound less terrifying or unfair.

Being useful and imaginative
This modification is in line with what people in the real world need since people who work in finance need to be able to make fair decisions based on facts. People won't have to guess or deceive if they know what the risks are.

You can also use the same prompt templates for more than one ESG concern, such ignoring labor standards or scandals in governance, or for multiple types of inquiry, like looking at ESG ratings. This suggests that it can grow and compare information.

You can also use the same prompt templates for more than one ESG issue, such scandals in governance or labor breaches, or for different sorts of research, like putting ESG ratings next to each other. This indicates it can get bigger.

The AI learning is very important in the changing landscape for all the team members as the management and Companies are gearing up to introduce the AI tools into their day to day Business processes.
The clarity with which the prompts are provided will be resulting in the more accurate answers from the systems.
The prompts need not be cluttered with too many queries clubbed together in one prompt as the AI models are defined and designed in a certain way to provide answers to queries through a prompt.

For example, say there is being AI used in a CRM process where team member is working on the lead generation and there are historical data available related to the leads which came and processed and then the leads which got generated through the leads.
The prompt provided by the Sales team member is as below:

"Find out similar timeline lead with similar value and similar OEM from where the enquiries got generated and got converted to Order and received the Sales Order"

Now the intent of the team member is to arrive at Final answer where the team member wants to know about the below and the prompts has to be sequential with more specific queries and should have been prompted as below:

First Prompt
i) Can you find a Lead from the OEM (name i.e. say Siemens, as there can be multiple OEMs i.e. Siemens, GE, Fanuk etc.) in the past?
Answer: Below mentioned Leads are from Siemens

Siemens - 2023 - 200 Cr. - 2 Years
Siemens - 2020 - 100 Cr. - 5 Years
Siemens - 2021 - 50 Cr. - 2 Years
Siemens - 2022 - 75 Cr. - 3 Years
Siemens - 2023 - 100 Cr. - 6 Years
Siemens - 1998 - 100 Cr. - 7 Years
Siemens - 2020 - 100 Cr. - 8 Years
Siemens - 2023 - 200 Cr. - 2.5 Years
Siemens - 2023 - 200 Cr. - 2.5 Years

Second Prompt
ii) Can you provide Lead with similar value?
Siemens - 2023 - 200 Cr. - 2 Years
Siemens - 2023 - 200 Cr. - 2.5 Years
Siemens - 2023 - 200 Cr. - 2.5 Years

Third Prompt
iii) Can you provide lead with similar timeline for the completion?
Siemens - 2023 - 200 Cr. - 2.5 Years
Siemens - 2023 - 200 Cr. - 2.5 Years

With the effective use of the clear, precise, sequential and specific prompt the right answers can be obtained from the AI model.
The right training and the with the practice the team members can get the right answer from the right prompt.
Prompt engineering needs to be learnt by the team members in order to use the AI to the fullest.
 

Ayomide Otokiti has provided the best answer to this question. Well Done!!

 

Response from Sumukha is also a good read!