Wireframe Your Generative AI Solution Design
Use Info-Tech's design methodology to align technical solutions with organizational needs.
- You recognize the ability of LLMs to revolutionize user experiences but struggle to design innovative solutions that leverage the technology.
- You don't have a formalized design methodology to integrate LLMs effectively into your solutions. You don’t have a systematic process to analyze LLM-integrated solutions.
- You don’t know how to balance technical innovation with organizational needs, and mismatches have hindered deployments or resulted in failed projects.
Our Advice
Critical Insight
Design solutions with a sociotechnical lens. Consider not only the technology but also the social systems within which the solution will be embedded to maximize its adoption, usefulness, and strategic alignment.
Impact and Result
- Use a human-tech ladder to systematically analyze requirements and constraints at multiple levels of need and align human need and technology outcomes.
- Identify challenges and unknowns during solution development and frame them as learning outcomes that scaffold concepts to be understood, applied, and leveraged for creative innovation.
- Leverage the solution library to understand Info-Tech’s design methodology while also exploring new concepts and ideas for generative AI technology.
Wireframe Your Generative AI Solution Design
Use Info-Tech’s design methodology to align technical solutions with organizational needs.
Analyst perspective
Generative AI technologies are opening the door to new types of solutions and user experiences. While exploring different ideas and concepts, it is useful to wireframe concepts to aid discussion with both technical and business stakeholders. For service providers, communicating ideas and documenting opportunities with customers should span social, human, and technical considerations. Ensure these areas are captured in early discussions to accelerate ideation and drive profitable projects forward.
Info-Tech’s methodology to wireframe generative AI solution designs will help you systematically consider organizational, team, user, model, data, software, and hardware elements in potential solutions. The methodology is a means to engage, align, and communicate with decision-makers, ensuring everyone is on the same page prior to making an investment.
Info-Tech’s solution library exemplifies its methodology through a deep dive into different solution designs. The solution library is intended to showcase some novel opportunities for solutions that leverage the unique capabilities of generative AI. You can use the library as a starting point to leverage the methodology or to kick off ideation with customers.
Info-Tech’s approach to wireframing a solution design is not one-size-fits-all. Rather, it is another tool for your toolbox that can help you generate value as a trusted technology service provider.
Dr. Justin St-Maurice
Principal Research Director
Technology Services
Info-Tech Research Group
Executive summary
Your Challenge |
Common Obstacles |
Info-Tech’s Approach |
|---|---|---|
You recognize the ability of generative AI to revolutionize user experiences but struggle to design innovative solutions that leverage the technology. You don't have a formalized solution design methodology to integrate generative AI effectively into your solutions or a systematic process to analyze requirements. You don’t know how to balance technical innovation with organizational needs, and mismatches have hindered deployments or resulted in failed projects. |
You don’t align solutions with specific end-user requirements or organizational contexts, creating a gap between technology delivery and user expectations. You don’t design with technical specifications and human requirements equally in mind and don’t know how to navigate complex sociotechnical design. You don’t focus on the human-technology intersection of solution design and struggle to simultaneously communicate requirements to business stakeholders and technology specialists. |
Align human needs and technology outcomes with the support of a human-tech ladder and systematically analyze requirements and constraints at multiple levels of need. Identify challenges and unknowns during solution development and frame them as learning outcomes that scaffold concepts to be understood, applied, and leveraged for creative innovation. Leverage Info-Tech’s solution library to understand the design methodology while exploring new concepts and ideas for generative AI technology. |
Info-Tech Insight
Design solutions with a sociotechnical lens. Consider not only the technology but also the social systems within which the solution will be embedded to maximize its adoption, usefulness, and strategic alignment.
Solutions leveraging generative AI can deliver experiences that were previously impossible
Unleash novel ideas with the unprecedented interactive capabilities of generative AI (Gen AI).
ChatGPT prompt [June 2024]: Draw me a picture of an LLM. Emphasize the chaotic nature of the LLM, its social interface and its advanced technology underpinnings. |
Previously impossible? Context-aware content generation: Gen AI produces rich, context-aware content adapted to cultural nuances, social contexts, and user needs. Brainstorming and decision-making support: Gen AI can analyze and synthesize huge amounts of data to provide insights and assessments that support brainstorming activities and decision-making. Personalization at scale: Gen AI can personalize user experiences in real time by catering to individual preferences, needs, and behaviors. |
Old technology, more parameters Large language models (LLMs) are closed systems of interconnected parameters. The most recent ancestor of LLM technology is simple autocorrect technology on cellphones. In the last decade, the technology has gone from taking small text inputs to predict the most appropriate word to come |
Gen AI increases solution complexity, increasing the need to systematically document requirements.
GEN AI SOLUTIONS ARE COMPLEX SOCIOTECHNICAL SYSTEMS.
Gen AI solutions are built on LLMs that combine algorithms, data, and infrastructure to dynamically interact with users. These models are powered by chaotic cultural norms and aggregated human knowledge.
LLMs bridge the boundary between AI and human behavior in unprecedented ways. With a virtually infinite number of input parameters, LLMs influence and are influenced by each interaction they have. The subjectivity of interactions introduces unique design constraints.
Technology service providers (TSPs) need to document and manage multiple interconnected requirements to successfully drive new experiences with Gen AI. The need to document requirements and align stakeholders grows exponentially alongside the complexity of the technology.
Poorly documented requirements are the number one cause of IT project failure.
(Statista, 2015)
It’s a new type of human/computer interaction
WHY IS INTERACTING WITH GEN AI SOLUTIONS CHALLENGING?
They produce complex sociotechnical impacts. Generally, information systems interact with people and impact multiple domains through complex interdependencies. Healthcare systems, for example, simultaneously impact data, patient care, physician collaboration, and regulatory requirements. Gen AI components further complicate these interactions.
They are dynamic, evolving pieces of technology. The underlying LLM models evolve with new data sets and will change with societal shifts in values, culture, and knowledge. At their core, they are not rule-based solutions and are difficult to manage with traditional technology controls.
They are nondeterministic systems. Gen AI is not designed to simply handle predetermined inputs. Its strength is its ability to accept input from any user and flexibly cater to different requests. Gen AI can infinitely customize experiences, and the integrity of such solutions requires new types of boundaries.
They need to incorporate responsible principles. While processing inputs, Gen AI solutions need to navigate ethical dilemmas and governance constraints. Data privacy, algorithmic bias, and the responsible use of AI are central to viable designs, where accountability becomes both a technical and a social challenge.
Review Get Started With AI in Requirements Management to learn more about the evolving human/computer interface.
Misaligning the human/computer experience is a recipe for failure
Misalignments between technological capabilities and social elements lead to failure, despite the intrinsic quality of the technology or the strategic soundness of the goals.
- Accurate models that act contrary to responsible principles will not be trusted.
- Poor data will degrade the user experience.
- Simple interfaces incorporated into complex workflows will hinder team productivity.
- Slow hardware that takes too much time to deliver a response will lower the value to the user.
BUY OR BUILD?
Misalignments between humans and technology occur in both buy and build scenarios. A sociotechnical systems lens is helpful in both approaches to designing integrated solutions. Selecting the right solution ensures that its implementation supports and advances the strategic objectives of the organization, regardless of whether the solution is engineered in-house or purchased off the shelf.
Embrace a sociotechnical systems approach to drive solution alignment
Social requirements are an important part of solution design. Sociotechnical systems (STS) theory is based on the idea that communities function through technology use, and that people work through software using hardware.
STSs are an intersection of multiple sciences:
- Sociologists see social systems
- Psychologists see cognitive systems
- Computer scientists see information systems
- Engineers see hardware systems
An STS approach sees multiple system types as interconnected and recognizes the union of people and technology as a natural evolution of both social and technical systems (Interaction Design Foundation, 2024).
Use an STS approach to:
- Explore the interrelationships between people, technology, and the organizational environment of solutions.
- Systematically align humans and technology to manage the interaction between them.
- Incorporate a balance between technically-advanced solutions and social responsibility by giving equal weight to human and technology elements.
Initial ChatGPT prompt [June 2024]: Draw me a picture of a socio-technical system.
Align human and technical elements using systems design principles and tools
Use Info-Tech’s methodology to align the human and technical elements of your innovative Gen AI solutions.
- Identify the human or organizational need.
- Decompose solutions requirements with human and technical lenses.
- Describe the human-technology interaction.
What is a design methodology?
A design pattern is a description or template for how to solve a software engineering problem in many different situations. A design pattern is generally narrowly focused and offers a detailed, reusable solution (Source Making, 2021).
A design methodology offers a structured framework that addresses broader system challenges and is adaptable to the unique requirements and complexities of many situations. A design methodology facilitates a systematic exploration of needs, design, and implementation through a holistic approach (IBM, 2023).
Using systems design methodologies to manage complex sociotechnical systems is a well-known and well-published domain of engineering science.
Info-Tech’s sociotechnical wireframe methodology for AI solutions
Design solutions with a sociotechnical lens.
Consider not only the technology but also the social systems within which the solution will be embedded to manage requirements and constraints.
Design Wireframe Methodology
Consider social and technical needs.
Describe the human or organizational need for the solution
Purpose
To develop a core understanding of the solution value from a human perspective and ensure the solution is rooted in addressing real, specific human or organizational needs.
Outputs
A problem statement that clearly and concisely describes the issue being addressed by the solution.
- Problem statements should be supported by concrete data.
- Problem statements should not prematurely suggest solutions.
Potential activities
- Interview users and key stakeholders.
- Distribute surveys to get insights on organizational needs and pain points.
- Analyze current systems and workflows to identify gaps and areas for improvement.
- Review social, technological, economic, and environmental factors that influence the need for the solution.
Develop a problem statement
Focus on understanding the problem fully before jumping into solutioning. In sociotechnical design, technology solutions need to address a specific human, social, or organizational need.
Why create a problem statement?
- It establishes a human-centric focus, emphasizing the user needs and contexts before defining technology requirements, and ensures solutions are tailored to actual human use cases.
- It mitigates technology-centric thinking and reduces the risk of developing solutions that are technically sound but fail to address real-world problems. Avoid technology for the sake of technology.
- It enhances solution adoption and increases the likelihood of user acceptance and satisfaction by aligning solution functionalities with specific user needs and expectations.
Clearly describe the problem
A problem statement should clearly and concisely describe the issue being addressed, outline the impact of the problem, and hint at possible outcomes without suggesting specific solutions prematurely.
Good problem statements |
Bad problem statements |
|---|---|
Specific: Clearly define the problem with specific data and insights and avoid vague descriptions. Use quantifiable metrics to highlight the severity and impact of the problem. |
Too broad: Don’t include ambiguous statements without clear direction for solution development. For example, "We need to improve our business" is too broad and lacks actionable specifics. |
Support your statement with data
A problem statement should include references to data and be grounded in facts.
Examples from different industries:
- Healthcare: “Hospital readmission rates for heart failure patients have increased by 20% over the last year, leading to higher healthcare costs and reduced patient satisfaction. A solution is needed to identify contributing factors and reduce these rates.”
- Education: “Local high school students are showing a 30% lower proficiency in math compared to national averages, affecting their readiness for college-level courses and future employment opportunities.”
- Technology: “Our company's software deployment process takes, on average, 30% longer than the industry benchmark, leading to delays in updates, decreased productivity, and client dissatisfaction.”
- Retail: “Customer foot traffic in our main store has declined by 15% quarter-over-quarter, significantly impacting sales and threatening long-term business sustainability
Specific, relevant, user-centric, and scoped
Example of a well-written problem statement for an LLM-based book recommendation solution at a public library:
| “Public libraries play a pivotal role in engaging communities with great and diverse literature, but with over 130 million books worldwide, finding a good book can be daunting. Advanced readers in particular face challenges in discovering new titles that resonate with their varied interests, especially readers from different cultural backgrounds. There is a need for an effective system within public libraries that can navigate the literary landscape without influence from marketplaces that can intelligently match readers with books that not only pique their interest but also expand their cultural understanding.” | ✓ Specific ✓ Relevant ✓ User-centric ✓ Scoped |
Use the human-tech ladder to define the solution
Purpose
To structure and define the solution in a way that aligns technological capabilities with human needs and organizational objectives, ensuring that the solution is both technologically feasible and human-centered.
Outputs
A solution with detailed requirements and constraints. Based on systematic analysis and according to the human-tech ladder lens, the solution:
- Provides an opportunity for input from a wide range of stakeholders.
- Defines a successful solution.
Activities
- Hold brainstorming sessions (e.g. whiteboarding sessions, tabletop exercises) with business and technical leaders.
- Use multiple iterations of activities to finalize the solution details based on input.
Leverage a human-tech ladder to define requirements
What is a human-tech ladder?
Vicente (2003) presented a human-tech ladder to describe technology uses at different levels of abstraction, including physical, psychological, team, organizational, and social levels. This methodology has been used in areas such as aviation, nuclear power generation, and healthcare to make technology safe for human interaction and useful for end users.
Info-Tech’s adaptation of the human-tech ladder decomposes AI solutions into their social and technological components to emphasize and manage the complexity of integrating AI systems within social and technical contexts. The modified ladder introduces layers specifically dedicated to AI, data, software, and hardware to help development a wholistic understanding of solution requirements involving AI.
The human-tech ladder can help you systematically consider both the social (human) and technical challenges at each stage of solution design and technology implementation to ensure a more holistic and integrated approach to developing complex sociotechnical systems.
Leverage the human-tech ladder to deliver great solutions
Why use the human-tech ladder?
Systematically assess potential challenges and specific requirements from multiple perspectives. Each level is a different type of challenge to manage and requires different competencies. Remember that managing people, policies, and organizational behavior is a different discipline than optimizing bits, bytes, and performance.
Simultaneously integrate solutions into multiple complex systems. Acknowledge and consider multiple impacts of the solution’s introduction and ensure that systems are aligned to drive an effective implementation.
Involve and engage multiple stakeholders and decision-makers in the design process and ensure that the needs and expectations of various stakeholders are met. Improve satisfaction among users, teams, and organizational leaders by designing solutions with their specific requirements in mind.
Systematically consider human and technical challenges
About Info-Tech
Info-Tech Research Group is the world’s fastest-growing information technology research and advisory company, proudly serving over 30,000 IT professionals.
We produce unbiased and highly relevant research to help CIOs and IT leaders make strategic, timely, and well-informed decisions. We partner closely with IT teams to provide everything they need, from actionable tools to analyst guidance, ensuring they deliver measurable results for their organizations.
What Is a Blueprint?
A blueprint is designed to be a roadmap, containing a methodology and the tools and templates you need to solve your IT problems.
Each blueprint can be accompanied by a Guided Implementation that provides you access to our world-class analysts to help you get through the project.
Talk to an Analyst
Our analyst calls are focused on helping our members use the research we produce, and our experts will guide you to successful project completion.
Book an Analyst Call on This Topic
You can start as early as tomorrow morning. Our analysts will explain the process during your first call.
Get Advice From a Subject Matter Expert
Each call will focus on explaining the material and helping you to plan your project, interpret and analyze the results of each project step, and set the direction for your next project step.
Unlock Sample ResearchContributors
- Peter Meyers, AI Advisory Lead, MSS Business Transformation Advisory
- Stephen Matusiak, Chief Technology Officer, HD Mutual Insurance