Introduction
In the process of elucidating ideas for applying Generative AI (GenAI) to solve business problems, establishing a series of workshops can be instrumental in refining these ideas and surfacing the best of the best. Workshops provide a structured environment for cross-functional teams to collaborate, brainstorm, and assess the feasibility and potential impact of various GenAI use cases. This collaborative approach not only fosters innovation but also ensures that diverse perspectives are considered, leading to more robust and well-rounded solutions.
Workshops play a crucial role in aligning stakeholders, setting clear objectives, and creating a shared understanding of GenAI's capabilities and limitations. By bringing together experts from different domains, these sessions can help bridge the gap between technical possibilities and business needs, ensuring that proposed solutions are both technically viable and strategically valuable. Moreover, the iterative nature of workshops allows for continuous refinement of ideas, enabling teams to adapt and improve their approaches based on feedback and new insights.
Assemble the Team: This is a Team Sport
Assembling the right working group is the cornerstone of successful GenAI workshops. This process should involve bringing together individuals with diverse expertise, including experts in the problem domain, data scientists, AI specialists, and business leaders. Each team member brings a unique perspective and set of skills that can contribute to the identification and development of innovative GenAI use cases. Ideally this would be done with the client directly or at the very least a surrogate with sufficient domain knowledge playing the role of the client. By fostering a collaborative environment, team members can share their knowledge, challenge assumptions, and build on each other's ideas, leading to more creative and effective solutions.
It's important to ensure that the team is not only diverse in terms of expertise but also representative of the various stakeholders who will be impacted by the GenAI applications. This inclusion helps to surface different needs and concerns, ensuring that the proposed solutions are comprehensive and address real-world problems. Furthermore, having a well-rounded team can facilitate better communication and alignment throughout the project lifecycle, from initial design ideas to final implementation.
A No-Fluff GenAI Primer: Don’t Give Me the Markitecture
A foundational step in any GenAI workshop is providing participants with a clear, no-nonsense understanding of what GenAI can and cannot do [obviously, this will need to be streamlined to the knowledge level of the participants and the various business challenges at hand]. This involves moving beyond the marketing hype and focusing on the actual capabilities, limitations, and practical applications of GenAI technologies. A well-structured primer should cover the basics of GenAI, its underlying mechanisms, and real-world examples of successful implementations [especially as aligned with the problem domain of the business]. This helps to ground the team in reality, setting realistic expectations and dispelling myths [no, GenAI will not do your laundry].
Understanding the technical nuances of GenAI is crucial for effective brainstorming and ideation. By demystifying the technology, participants can better identify feasible use cases and avoid pursuing ideas that are technically impractical or overly ambitious. This step also empowers non-technical stakeholders to contribute more effectively to the discussion, ensuring that their insights and needs are integrated into the development process.
Brainstorm Use Cases: Fostering Creativity with Strategic Insight
After the GenAI primer, the next crucial step for the working group is brainstorming potential use cases. This stage is where creativity meets strategy, allowing the team to think expansively about how GenAI can be applied to solve real business problems. Brainstorming sessions should be inclusive and collaborative, encouraging input from all team members regardless of their role or seniority. This diversity of perspectives can unearth unique and innovative ideas that might otherwise be overlooked. The key is to create an environment where participants feel free to share their thoughts without fear of criticism, fostering a sense of collective problem-solving. I would also add that it is probably a good idea to “seed” the conversation with a few sample use cases of leveraging AI in the problem domain with a simplified workflow; this may help get some participants to lean into the “fray”, so to speak.
During brainstorming, it’s essential to keep the prioritization attributes in mind. While the goal is to generate a wide array of ideas, having a framework for evaluation helps ensure that the brainstorming is focused and aligned with the organization’s strategic objectives. Attributes such as business impact, feasibility, innovation potential, and regulatory compliance should be guiding factors in this creative process. By considering these attributes from the outset, the team can more effectively filter and refine ideas, making the subsequent prioritization step more streamlined and grounded in strategic relevance.
Incorporating prioritization attributes into the brainstorming session also helps set realistic expectations and directs creative energy toward practical solutions. It encourages team members to think not just about what is possible, but what is impactful and achievable. This approach bridges the gap between ideation and implementation, ensuring that the ideas generated are not only innovative but also actionable and aligned with the organization’s capabilities and goals.
Use Case Evaluation/Prioritization: Let’s Pick Our Best Shot
Once the team is assembled and equipped with a solid understanding of GenAI and have a few use cases in mind, the next step is to evaluate and prioritize potential use cases. This involves identifying key attributes and criteria that will guide the development and prioritization of each use case. Factors such as business impact, feasibility, regulatory compliance, and innovation potential should be considered to ensure that the most promising and valuable use cases are selected for further development.
I have an entire blog post on this topic with ideas on key attributes/considerations:
Establishing a clear and objective framework for evaluating use cases helps to ensure that decisions are data-driven, aligned with strategic goals, and feasible. This process should be collaborative, with input from all team members to capture a wide range of perspectives and insights. By systematically assessing each use case against predefined criteria, the team can prioritize projects that offer the greatest potential for success and impact.
I should add that this step in the workshop cycle would probably need to be iterated on such that the important details of the use cases are known and agreed among the participants (before the “shark tank”).
Bringing the Ideas to The Shark Tank
The final step in the workshop process is presenting the prioritized ideas to senior business leaders for vetting and approval. This "Shark Tank" approach involves pitching the refined use cases to decision-makers who can provide valuable feedback, identify potential challenges, and ultimately decide which projects should move forward. Engaging senior leadership early in the process ensures that the selected use cases are aligned with organizational priorities and have the necessary support and resources for successful implementation.
I really like this step in the workshop process because not only is it critical for securing buy-in and commitment from key stakeholders, but they can open your eyes to big holes in the tone/tenor (etc.) of the pitch. By involving senior leaders in the evaluation process, teams can gain insights into strategic priorities and constraints that may not have been considered previously. Additionally, this approach fosters accountability and transparency, as decision-makers are directly involved in the selection and prioritization of GenAI projects, ensuring that the chosen initiatives are well-supported and poised for success.
Additional Steps/Elements for GenAI Application Development Workshops
Develop Initial Prototypes: Bringing Ideas to Life
Once the best ideas have been identified and prioritized, the next step is to develop initial prototypes or proof-of-concept (POC) models. This step involves small, cross-functional teams working on building rudimentary versions of the selected use cases. The aim is to test the feasibility and potential impact of the ideas in a controlled environment. Prototyping allows teams to quickly iterate and refine ideas, making necessary adjustments based on early feedback and performance metrics.
Further, where client data is scarce or needs to be protected, you can always consider using Generative Adversarial Networks (GAN) to generate realistic data for your prototype. I’m planning a future post on GANs.
Feedback and Iteration: Refining the Prototypes
After the initial prototypes are developed, gather feedback from a diverse group of stakeholders, including end-users, business leaders, and technical experts. This feedback is critical for identifying strengths, weaknesses, and areas for improvement. The iterative process of refining the prototypes based on real-world feedback helps ensure that the final solutions are robust, user-friendly, and capable of delivering the intended business value. This step emphasizes the importance of continuous improvement and adaptability in the development of GenAI applications.