5 Years of AI Design Sprint: Transforming How Businesses Approach Artificial Intelligence

I’ve spent countless hours immersed in the world of workshops- preparing agendas late into the night, rushing to the print shop to pick up freshly designed canvases, and standing in front of teams with a marker in hand, ready to guide them through big, messy challenges. Whether it was aligning stakeholders, brainstorming solutions, or diving deep into Design Thinking, I loved the energy of bringing people together to solve problems.

As an AI Product Lead, my journey with AI Design Sprints has been a fascinating evolution. Working at nexocode, a company deeply focused on data science, machine learning, and artificial intelligence, revealed how the complexity of AI technologies changes not just project implementation but also how we define and manage them.

It quickly became clear that clients seeking AI solutions often struggled with more than just execution. They needed help aligning objectives, defining realistic goals, and ensuring their teams had a comprehensive understanding of AI’s capabilities and limitations. Integrating AI into their business intelligence strategies was crucial for driving business value. This gap inspired us to rethink the traditional workshop model and tailor it to meet these emerging needs.

Our AI Design Sprints were designed to simplify the complexities of artificial intelligence, empower teams with actionable insights, and foster AI knowledge-sharing for business leaders and client teams alike.

Five years later (wow!), I’ve facilitated more workshops than I can count- working across industries, learning about challenges I never even knew existed, and collaborating with incredible people. We’ve built dozens of specialized exercises, adapted to groundbreaking technologies like Generative AI, and helped businesses navigate the journey to smarter AI adoption. Along the way, we’ve learned a ton - about AI, about people, and about what it really takes to bring those two worlds together.

This article is my chance to reflect on that journey, share the lessons we’ve learned, and explore what’s next for this methodology.

What is AI Design Sprint?

The AI Design Sprint is a structured approach to developing innovative AI solutions that drive business value. It is a collaborative process that brings together cross-functional teams, stakeholders, and AI experts to identify, design, and prototype AI-powered solutions. The AI Design Sprint is designed to help organizations overcome the challenges of AI adoption and unlock the full potential of artificial intelligence.

In essence, the AI Design Sprint is about bridging the gap between ambitious AI ideas and practical, actionable outcomes. By fostering a collaborative environment, it ensures that everyone—from business leaders to technical experts—works together to explore AI capabilities and develop solutions that align with business goals. This structured approach not only accelerates the development of AI solutions but also enhances the overall understanding of artificial intelligence within the organization.

Why We Needed a Dedicated Format for AI and Data Projects

Before we developed the AI Design Sprint, I often relied on traditional Design Sprints to help teams tackle complex challenges. These workshops were effective in many contexts, but as our work at nexocode shifted more toward data science, machine learning, and artificial intelligence, it became clear that something wasn’t clicking. AI projects weren’t like other projects—they required a deeper understanding of data, business intelligence, the right AI tools, and a clear focus on aligning business goals with technical possibilities.

Why AI Projects Are Different

Working on AI projects is nothing like building traditional software. In software development, you usually start with a clear set of requirements, and the logic is straightforward—if you follow the plan, you know what to expect at the end. AI, though, is a whole different story. It’s messy, unpredictable, and comes with its own set of risks and costs.

One of the biggest differences is that many AI solutions feel more like science experiments than engineering projects. When you’re building a traditional feature, you can see the path ahead. With AI, you’re experimenting, trying to see if a machine learning model can do what you hope it can. And here’s the catch: you often don’t know until you try.

For instance, I’ve seen models fail simply because the data wasn’t up to the task—maybe it wasn’t diverse enough, didn’t have the volume needed, or wasn’t of high enough quality. That’s where the uncertainty comes in. You can put in all the effort, but if the data doesn’t support the insights you’re looking for, the project might not succeed as planned.

I remember one of our early projects—a client approached us with big plans for an AI solution. They envisioned deep learning systems transforming their business processes but couldn’t clearly define the problem they wanted to solve or the data they needed. That was a lightbulb moment for me: the standard Design Sprint wasn’t enough for AI and emerging technologies.

We needed to address gaps in how businesses were applying AI. It wasn’t just about building the AI model; it was about helping teams gain insights into their data, setting realistic expectations, and fostering a comprehensive understanding of AI’s potential. This realization led us to develop the AI Design Sprint—a specialized format tailored for the complexities of AI adoption and data-driven development.

With this approach, we defined what we need to focus on:

• Helping teams define objectives Many clients come with ambitious ideas but no clear business opportunities tied to them. With all the hype on AI comes the misconception that projects are simple, magical solutions that can solve any problem at the push of a button. In reality, they are often far more complex than expected—deeply specific, grounded in mathematics and statistics, and highly dependent on quality data and careful planning. The AI Design Sprint ensures teams identify specific problems and set measurable outcomes for their AI applications.

• Focusing on model outputs One of the most common misconceptions we encounter is the belief that an AI model will act as a magical black box that solves all problems at once. To address this, we created exercises that guide teams in defining tangible model outputs. For example:
• A recommender model outputs a list of items matching a selected product.
• A pricing model dynamically calculates the price for a specific cargo definition.
• A fraud detection model flags potentially fraudulent claims.
  Super important exercise, and honestly not that easy to run. But by narrowing down outputs to specific, actionable results, we help clients set realistic expectations and create a clear path toward success.
• Understanding data One thing we quickly learned is that no AI project succeeds without good data. Our workshops include dedicated exercises to identify the necessary datasets, evaluate their quality, and develop a roadmap for better data processing.
• Simplifying AI technologies Let’s face it—terms like deep time series clustering, generative AI, or even AI agents can feel overwhelming to non-technical stakeholders. These workshops make AI accessible, breaking down complex AI systems into practical concepts that teams can rally behind.
• Encouraging collaboration and learning It’s not just about creating solutions—it’s about empowering teams. By fostering a collaborative approach, we turn our sprints into spaces for learning experiences, where business leaders and technical teams can align around the possibilities and limitations of AI.

Over the past five years, I’ve seen firsthand how this shift has impacted our clients. For some, the AI Design Sprint was the spark they needed to kickstart their AI adoption journey. For others, it was an eye-opener that led them to first build a stronger data strategy or explore smarter AI tools before diving into ambitious AI solutions.

Lessons Learned: How AI Design Sprints Have Evolved Over 5 Years

When we first started running AI Design Sprints, we were focused on adapting the Design Sprint methodology to fit the unique challenges of AI technologies. Five years later, what we have today is a robust and dynamic framework built from experience, iteration, and countless client engagements. The journey has taught us invaluable lessons - not just about artificial intelligence but also about how teams and businesses navigate the complexities of AI adoption.

Evolving the Toolkit

One of the most exciting parts of this journey has been watching our toolkit grow to help teams understand and apply AI in their business intelligence strategies. What began as a handful of exercises has transformed into a diverse collection of resources, tailored for specific industries, project types, and technologies.

• We now have over 50 specialized canvases, each designed to guide teams through critical aspects of an AI project—from data preparation to defining AI model outputs.
• We developed four sets of card decks, helping participants quickly grasp AI concepts, challenges, and use cases in an engaging, hands-on way.
• To make the workshops more interactive, we introduced playgrounds where teams can explore AI tools, experiment with mock data, and see potential solutions in action.
• Recognizing the need for long-term planning, we created calculators that help clients evaluate the costs, ROI, and maintenance efforts associated with their AI projects.

Adapting to Remote and Hybrid Work

When COVID-19 hit, we had to completely rethink how we ran our workshops. Moving online was a challenge at first, but it also pushed us to innovate. We became power users of tools like Miro, reimagining the workshop experience to keep it just as engaging and collaborative as in-person sessions.

Now, we’ve fully embraced a flexible offering that includes onsite, remote, and hybrid formats. This adaptability has allowed us to support clients wherever they are, without compromising the quality of the experience.

Bridging the Knowledge Gap with AI Training

One of the most valuable insights we’ve gained over the years is that many workshop participants, especially business leaders, struggle to fully grasp the nuances of artificial intelligence necessary to successfully run AI projects within their org. Terms like machine learning, deep learning, Generative AI, and AI model outputs can feel intimidating or overly technical, creating a barrier to effective decision-making and collaboration.

To address this, we expanded our AI Design Sprint framework to include dedicated AI training modules. These sessions are designed to demystify the technology, break down complex concepts, and provide participants with a comprehensive understanding of how AI works and what it can achieve. The training modules also enhance participants’ insights into applying AI in their business intelligence strategies, ultimately driving business value.

We use case studies and success stories from various industries (and of course client’s industry) to show how AI solutions can be applied to real problems. Using AI tools and interactive playgrounds, participants get hands-on training and feel for how models are trained, how they generate outputs, and where potential challenges might arise.

The Generative AI Impact

The rise of Generative AI marked a pivotal moment for us. Suddenly, there was a wave of excitement (and sometimes unrealistic expectations) around AI solutions. Clients came to us with ideas that sounded impressive on paper but were often too complex or lacked clear objectives.

Through the AI Design Sprint, we learned how to break these ideas down into simpler, actionable projects. For example:

• Instead of building a catch-all AI-powered chatbot, we guided a client to focus on an AI agent tailored to automate a specific customer support process.
• Rather than attempting to revolutionize an entire supply chain, we helped narrow the project to a machine learning model for dynamic pricing of high-value cargo.
• For another client, what started as an ambitious plan for a predictive analytics platform became a fraud detection model to flag high-risk claims, an achievable, high-impact solution.

The Power of Stepping Back

One of the things that surprised me the most was how often the AI Design Sprint became a chance for clients to take a step back and rethink their data strategy.

Many realized their data wasn’t quite ready for AI, maybe it wasn’t clean enough, organized well, or suited for the kind of data science projects they had in mind. Instead of diving straight into building solutions, the sprint became about laying the groundwork for future success.

This shift in perspective was a game-changer. Some clients put their AI plans on hold to focus on cleaning and structuring their data and getting ready with their data architecture, while others completely redefined their approach, using the sprint as part of a broader strategy to reshape their organization and prepare for AI adoption the right way. It was a reminder that sometimes, slowing down is the fastest way to get ahead.

Tailoring to Every Level of AI Maturity

One of the biggest lessons we’ve learned is that no two organizations are the same when it comes to AI adoption. Startups, for example, often benefit from condensed, high-energy sprints focused on specific AI applications. Larger enterprises, on the other hand, may need the sprint to be part of a broader organizational change management strategy. AI workshops help businesses identify and prioritize scenarios where AI can provide a competitive advantage.

We’ve also learned to adapt our offering based on the problem space. If a client has already evaluated their challenge, we suggest a more intense, fast-paced sprint. For organizations just starting their AI journey, the sprint becomes a mix of strategy, learning experiences, and knowledge sharing.

Lessons About Teams and Collaboration

One thing that’s been reinforced again and again is the importance of alignment. AI projects often fail not because the technology isn’t there but because teams aren’t on the same page. The AI Design Sprint creates a space for collaboration, where technical teams, business leaders, and decision-makers come together to align on goals, expectations, and roles.

The Cost of AI: Why Calculating ROI Matters

One of the most eye-opening lessons I’ve learned is how often businesses underestimate the cost of building AI solutions. It’s easy to see tools like ChatGPT and think, “If this is available for free (or at a low subscription price), surely we can have something just as big for a small budget.” But the reality is far from that…

Most AI projects require significant upfront investments. There’s the data preparation, which is often far more time-consuming than anyone expects, and then model training, which can be GPU resource-intensive. And it doesn’t stop there- there are ongoing costs like model inference, cloud infrastructure, and the inevitable model retraining to keep things accurate as conditions or data evolve.

I’ve seen clients come into a sprint with grand ideas, only to realize that the financial realities of AI mean they need to think carefully about ROI. That’s why we’ve made ROI discussions a core part of the AI Design Sprint. These conversations aren’t always easy, but they’re necessary. We work through exercises to help teams map out the costs versus the value their AI system could deliver- whether it’s saving time, improving efficiency, or driving new revenue streams.

I’ve found that this process can be a real turning point for businesses. For some, it’s a reality check, they realize they need to invest more time in planning or refining their ideas before jumping into development. For others, it’s a moment of clarity that gives them confidence in moving forward because they understand both the challenges and the opportunities.

What’s been most rewarding for me is helping teams make smarter decisions. AI isn’t magic, and it’s not about chasing trends, it’s about making thoughtful, strategic investments that pay off in the long run. Helping clients see the bigger picture and plan for success has been one of the most valuable outcomes of the work we do.

Tips for Businesses Adopting Design Thinking Tools in AI Solution Implementation Projects

Implementing an AI solution can be daunting, but design thinking tools offer a practical and collaborative way to navigate the complexity. From aligning teams to defining clear objectives, these tools provide structure and clarity at every stage of an AI adoption journey. If you’re looking to bring similar tools into your internal processes, here’s what I’ve learned:

1. Start with problem definition Before jumping into solutions, use tools like ideation canvases or our AI Implementation Kickstarter Canvas (free to download) to define the problem clearly. AI thrives on specificity-understanding the business challenge you’re solving is critical to success. Ask questions like, “How might we improve x process that currently takes a lot of our time and is prone to error?” or “What tangible outcomes do we expect from this AI application?”
   
2. Focus on collaboration
   AI projects often involve diverse teams with varying levels of technical expertise. Tools like Miro or collaborative whiteboards can help bridge the gap, ensuring everyone has a voice and a shared understanding. The visual nature of these tools helps align teams quickly, especially when tackling complex concepts like AI capabilities or business processes. Use visual tools to map out workflows, data requirements, and potential bottlenecks in the process.
   
3. Embrace hands-on learning
   Another powerful approach is to look for case studies and success stories—whether from your own industry or from similar projects applied in completely different fields. These examples offer a wealth of inspiration and practical insights, showing what’s possible while also highlighting potential challenges to anticipate. Seeing how others have successfully implemented AI solutions can spark ideas and give teams the confidence to adapt those concepts to their unique needs. Plus, studying successes (and failures) helps ground ambitious ideas in reality, aligning innovation with achievable outcomes.
4. Balance innovation with practicality
   It’s easy to get caught up in the excitement of emerging technology, but successful AI projects require a balance between ambition and feasibility. Design thinking tools help teams stay focused on delivering tangible outcomes without over-complicating the solution. Keep asking, “What’s the simplest version of this that will deliver value?” Tools like calculators or canvases are invaluable for grounding ambitious AI projects in reality. You can use them to clarify costs, benefits, and next steps- it’s a great way to foster confidence and avoid over-engineering.

Advice for Businesses Starting Their AI Journey

Starting your AI journey can feel overwhelming, especially with the abundance of possibilities and the hype surrounding artificial intelligence. But over the years, I’ve learned that taking a thoughtful, measured approach makes all the difference. Here’s my advice for businesses just getting started:

1. Be clear on what you want to achieve

Before jumping into AI, take a step back and ask yourself: “What problem are we trying to solve?” AI isn’t a one-size-fits-all solution, and without a clear objective, it’s easy to lose focus. Defining business challenges and opportunities for AI adoption is crucial for integrating AI into business intelligence strategies. Key questions to consider:

• What specific business challenge are we addressing?
• How will we measure success?
• Are there simpler, non-AI ways to solve this problem?

Having a clear purpose will help guide every decision, from selecting the right AI model to defining measurable outcomes.

2. Build the right team

AI projects require collaboration across multiple disciplines, from data scientists and engineers to business leaders and process owners. Ensure your team has a mix of technical expertise and business acumen. If your organization doesn’t have in-house AI expertise, consider partnering with an experienced team to guide you. The right AI workshop or training can also help align and upskill your team.

3. Don’t underestimate the preparation phase

AI relies heavily on data, and preparing that data can be one of the most time-consuming parts of a project. Make sure your data is accessible, relevant, and clean enough to support your goals.

Data is undeniably critical for AI, but I’ve seen teams stall their projects entirely while waiting for their data to be “perfect.” In reality, no dataset is ever perfect. The sprint includes strategies to work with the data you have now while planning for data quality improvements as the project progresses. It’s about moving forward without getting bogged down.

Rather than waiting for perfection, start with the data you have and plan for iterative improvements. The process itself will often reveal gaps or areas for refinement.

4. Start small and scale later

It’s tempting to aim for big, transformative solutions, but the best way to approach AI is often to start small.

Some teams become obsessed with making their AI models flawless before they’ve even tested them in the real world. While accuracy is important, over-engineering can delay progress and inflate costs. We encourage teams to focus on building a minimum viable model- something functional enough to test and iterate upon.

Focus on solving one problem well before expanding to other areas. Early wins will build momentum and confidence within your organization.

5. Plan for post-workshop action

A successful AI Design Sprint is just the beginning. To see tangible results, you’ll need a clear plan for what happens after the workshop ends. Define next steps, assign responsibilities, and ensure everyone knows their role in moving the project forward.

6. Embrace a culture of learning and experimentation

AI projects are inherently experimental. Success often comes from trying, failing, and learning along the way. Encourage your team to view setbacks as opportunities for growth rather than failures. Incorporate time for reflection and feedback throughout the project lifecycle. Regular check-ins help ensure alignment and continuous improvement.

7. Seek inspiration from others

You don’t have to reinvent the wheel. Look for case studies and success stories from other industries or similar projects. These can spark ideas, provide practical insights, and help set realistic expectations. Adapt what works for others to fit your unique needs, but don’t be afraid to innovate and customize.

Starting your AI journey can feel like stepping into the unknown, but with the right mindset, tools, and preparation, it’s an incredibly rewarding process.

Remember, AI isn’t just about technology- it’s about solving real problems, driving meaningful outcomes, and unlocking new business opportunities. Taking that first step thoughtfully sets the stage for long-term success.

What’s in Store for AI Design Sprints in 2025?

As we enter 2025, the world of artificial intelligence is more exciting than ever and so are the opportunities and challenges for businesses looking to get ahead. This year we’re not just continuing the AI Design Sprints – we’re evolving them to stay ahead of the curve and deliver more impact.

Emerging technologies One of the biggest trends we’re seeing is the rapid progress in Generative AI, AI agents and foundation models. These technologies are opening up new possibilities for automation, creativity and decision making across industries. So we’re adding new exercises and tools to our sprint framework to help teams explore AI in these areas. Whether it’s how generative models can re-define customer engagement or deploying AI agents to simplify operations, we’re making sure our clients are equipped to innovate with confidence.

In 2025 we’re adding more tools to make complex AI concepts more accessible and actionable. We’re building more interactive playgrounds where teams can play with AI tools and prototype ideas in real-time. We’re also adding more canvases and templates to address emerging challenges – from ethical AI to sustainable AI adoption.

Flexibility is still core to our approach. As organizations move to hybrid work environments we’re refining our ability to deliver AI workshops online and onsite. We’re also exploring shorter, more focused sprints for teams with well-defined problems and longer programmes that integrate AI Design Sprints into broader organizational transformation initiatives.

The pace of AI innovation isn’t slowing down and neither are we. Our focus is on being agile – continuously refining the sprint framework based on client feedback, industry trends and advancements in AI technologies. This means being curious, listening to what businesses need and ensuring the AI Design Sprint evolves with the challenges and opportunities of the AI world.

Ultimately for 2025, we want the AI Design Sprint to be an even more powerful tool for businesses to not only adopt AI but to thrive in an AI world. By combining creativity, strategy and the latest technology we’re helping teams unlock the power of AI and turn their ambitions into reality.

Here’s to a year of new possibilities and turning ideas into success stories. Cheers! 🎉️👋️🔥️🎊️👍

Ready to get started?

If you’re looking to turn your AI ideas into actionable outcomes or need guidance on navigating the complexities of artificial intelligence, an AI Design Sprint could be exactly what you need. Whether you’re just starting your AI journey or looking to refine your strategy, we’re here to help.

Book an AI Design Sprint today or schedule a consultation to explore how AI can unlock new possibilities for your business.