Harnessing Business Analysis Skills for Generative AI Prompt Design

Jul 10

Introduction to Generative AI

Generative AI models, like OpenAI’s GPT and Google Bard, are trained to understand and generate natural language text based on user prompts. Influencing these models using prompts is like directing a conversation: if you ask a generative AI model a specific question, it will provide a specific answer, following the patterns and rules it learned from its training on vast amounts of content.

Business Analyst Skills and AI

Business analysts have a unique blend of skills well-suited to designing effective generative AI prompts, including requirements management, stakeholder analysis, and solution assessment. They have a unique opportunity to apply their skills to generative AI.

Understanding requirements translates to knowing what a prompt should accomplish.
Knowing the audience aids in customizing the prompts to the user's context.
Planning ensures that the prompts align with the broader objectives of the project.

Creating Prompts Requirements Management

The cornerstone of creating effective prompts lies in identifying and defining the right requirements, the primary focus of business analysis. Requirements development skills can transfer to AI prompt design.

For instance, if an AI project aims to assist customer service, the requirements might include understanding product information, handling basic customer queries, and escalating complex issues. The business analyst would then design prompts to elicit responses or generate content that aligns with these requirements.

Stakeholder Analysis Understanding the Audience

Knowing your audience is just as critical in AI as in business analysis. Stakeholder analysis aids in understanding the AI users’ context, and this knowledge is crucial in creating prompts that resonate with the audience.

For example, a customer service AI would interact with clients quite differently from service staff and management. Understanding the stakeholders' needs, the customers, service agents, and managers can help create empathetic and contextually appropriate prompts.

Solution Assessment Evaluating the Prompts

Creating prompts includes evaluating their effectiveness, thus bringing solution assessment into play. Business analysts are well-versed in measuring the success of a solution against its initial requirements. In the AI space, this might mean reviewing the responses generated by the AI to certain prompts, getting feedback from users, and making necessary adjustments to the prompts or even the AI model itself to ensure it meets the project's needs.

Customer Service Prompt Example

Here is an example prompt to create requirements for a responsive customer service solution. The prompt starts by defining the role ChatGPT will play and a concise summary of its task:

As a software development business analyst, create a list of business requirements for a customer service solution.

Next, the prompt specifies what’s needed from the solution; more specifically:

The solution should enable customers to quickly resolve their issues with a product or service. If a customer cannot find what they need, have a chatbot request additional information without asking for information they previously submitted. The bot should continuously gauge the customer's sentiment and escalate to a live representative if the sentiment turns negative. The representative should have background information on the customer as well as all information collected during the service request.

Finally, the prompt specifies the format of the results and directs ChatGPT to ask questions for clarification:

Present the requirements as a bulleted list. Ask me if you have any questions about this request.

ChatGPT’s response leans more towards a solution, probably because its training includes more solution-biased content:

1. Efficient Issue Resolution:

Enable customers to resolve their issues quickly and effectively.
Provide accurate and relevant information to address customer queries and concerns.
Minimize the need for customer escalation by offering comprehensive self-service options.

2. Intelligent Chatbot:

Implement a chatbot with advanced natural language processing capabilities.
Enable the chatbot to understand and respond to a wide range of customer queries.
Utilize machine learning algorithms to continuously improve the chatbot's performance and accuracy.

3. Contextual Information Gathering:

Design the chatbot to request additional information only when necessary.
Avoid repetitive information requests by leveraging previously submitted data.
Streamline the information collection process to minimize customer effort and frustration.

4. Sentiment Analysis and Escalation:

Implement sentiment analysis algorithms to continuously gauge customer sentiment.
Detect negative sentiment and dissatisfaction in real-time.
Automatically escalate to a live representative when negative sentiment is detected.

5. Comprehensive Customer Profiles:

Capture and store detailed customer information, including contact details and preferences.
Maintain a complete history of customer interactions and service requests.
Enable representatives to access customer profiles and interaction history for personalized support.

6. Integration with Backend Systems:

Integrate with customer relationship management (CRM) systems to access customer data.
Sync customer service interactions with CRM for a unified view of customer interactions.
Enable seamless information sharing between the customer service solution and other relevant systems.

7. Performance Tracking and Reporting:

Track key performance metrics, such as response time, first contact resolution rate, and customer satisfaction ratings.
Generate comprehensive reports and analytics dashboards to monitor and improve customer service performance.
Identify trends, bottlenecks, and areas for improvement through data-driven insights.

8. Multichannel Support:

Provide support across multiple channels, including web chat, mobile apps, and social media platforms.
Ensure a consistent and seamless customer experience across all channels.
Integrate with existing communication channels and systems to centralize customer interactions.

Please let me know if you have any further questions or if there's anything else I can assist you with!

ChatGPT concludes with an opportunity to drill down into its requirements. For example, a business analyst could ask:

How do we measure customers’ “resolving their issues quickly and effectively?”
How can we “Streamline the information collection process to minimize customer effort and frustration?”
Elaborate more on “Multichannel Support.”

Also, ”Integration with Backend Systems” suggests Einstein GPT, even though ChatGPT’s training preceded EinsteinGPT’s announcement.

The Future of Business Analysis and AI

The intersection between business analysis and AI burgeons with potential. As generative AI continues to evolve and influence business operations, the role of business analysts will expand and adapt. As a business analyst, understanding AI, and applying your unique skills to it, will not only enhance your toolkit but also future-proof your career in an increasingly dynamic world.
Generative AI prompt design is a new frontier with abundant opportunities for business analysts to leverage their talents and showcase their value. Embrace the potential of AI, dive into the nuances of prompt design, and harness the power of your business analysis skills to shape the future of AI in the business realm.

Harnessing the unique skills of business analysts in the realm of generative AI prompt design not only illuminates the immense potential in this novel intersection but also paves the way for an enhanced, future-proof career in the ever-evolving digital landscape.

Richard Cunningham