The Silent Killer - Do AI Projects Need a New Approach to Project Management?

Working on AI projects can be like trying to solve a puzzle when the pieces keep changing shape. Plans rarely stay solid for long because new data and discoveries shift the goalposts. AI product development is inherently complex and uncertain, with requirements and challenges frequently evolving as new insights and data emerge. That’s where traditional project management hits a wall. Classical methods like waterfall rely on a linear, sequential process that assumes requirements can be fully defined upfront - an assumption that doesn’t hold in complex AI projects.

What works better is a mindset built for unpredictability - one that lets teams adapt as they go. To navigate this, AI teams must adopt delivery-focused Agile frameworks emphasising rapid, iterative progress and adaptability. Complexity thinking is crucial here, as it acknowledges the unpredictable nature of AI projects and the need for flexibility to respond to shifting demands. Agile isn’t about rigid schedules - it’s about taking small steps, checking if you’re on the right track, and adjusting before things go too far wrong.

Through Agile’s learning feedback loops, teams can validate ideas quickly, refine models, and correct errors, reducing wasted effort and maintaining alignment with end-user needs. This kind of approach keeps teams focused on what actually matters and lets them learn and improve along the way. This iterative approach ensures that AI solutions evolve in response to real-world feedback, allowing for continuous improvement and enabling teams to deliver valuable products efficiently and safely. It’s not just a nice-to-have; for AI, it’s essential.

Why Leaders Can’t Avoid The Reality

Today's successful companies aren’t just experimenting with AI - they’re upskilling on how to manage it. They understand that there’s a way of mastering AI to ensure it performs in the best way possible. So, if you’re still clinging to rigid project management methods in this context, you’re already falling behind. Agile is an?essential skill?that companies must master to stay relevant in an ever-competitive world. You need to consider what could be lost if you’re not proactive in this space, such as:

• Missed opportunities: While you’re waiting for a perfect plan, your competitors are launching, learning, and improving.
• Time to market: They’re getting their solutions out faster, adapting in real-time, and scooping up market share.
• Cost of inaction: Delays and inefficiencies don’t just slow you down, they hit your bottom line hard too.

As Amazon founder Jeff Bezos said, “If you double the number of experiments you do per year, you’re going to double your inventiveness.” Agile gives you the tools to move fast, experiment more, and learn quicker than ever.

The fact is, if you’re managing AI projects, you need an approach that keeps up with the constant changes. Agile helps you stay on top of things and keeps everything moving smoothly. So leaders need to ask themselves whether they want to be the ones catching up or setting the pace.

Why Traditional Methods Fall Short

Conventional project management relies heavily on setting clear deliverables and sticking to a plan. While this works for predictable tasks, it can cause problems in AI initiatives, here’s why:

• Outcomes are uncertain: You won’t know if an AI model will work as expected until you’ve tested it thoroughly. Sometimes, it takes several iterations to get there.
• Frequent iteration is necessary: AI projects aren’t linear, instead they evolve through repeated cycles of building, testing, and refining.
• Team collaboration is key: Success relies on input from a range of experts, such as data scientists, engineers, and business leaders. Without consistent communication, progress can grind to a halt.

Relying on rigid plans in project management for AI projects is a bit like running a marathon in the wrong shoes - you’ll make progress, but it will be delayed and very uncomfortable! Flexibility is what keeps the project on track.

Why Agile Works

Agile flourishes in situations where flexibility and speed are crucial. It focuses on delivering value in small increments, allowing teams to learn and adapt as they go. Here’s why this approach works so well for managing AI projects:

• Flexibility to adapt: Plans in AI projects may need to change quickly. Maybe new data reveals a better path forward, or early testing highlights an issue. Agile helps teams adjust effectively without losing momentum.
• Rapid feedback cycles: Agile breaks work into short sprints, where teams deliver tangible progress in small, functional pieces. This makes it easier to spot issues early and make real-time adjustments. The sooner you can test and refine, the faster you’ll reach your goal.
• Encourages team collaboration: AI projects don’t happen in isolation. You need data scientists to build models, engineers to deploy them, and business leaders to guide the strategy. Agile is all about open communication, ensuring everyone is aligned and working towards the same objectives.

For teams managing AI projects, Agile is the perfect fit. It lets you test, tweak, and keep things moving without getting bogged down.

What Leaders Can Do Now

Adopting Agile for AI projects doesn’t require an overhaul overnight. Start small and build from there. Here are a few practical steps to get started:

• Break projects into manageable parts: Focus on delivering smaller pieces of value that can be tested and improved over time.
• Establish clear goals: Keep the team together by ensuring everyone understands the project’s broader objectives.
• Prioritise collaboration: Regular check-ins and open discussions can help maintain momentum and keep the project on track.
• Embrace a learning mindset: AI is evolving all the time, and not every attempt will succeed on the first try. So, use setbacks as opportunities to learn and improve.

The complexities of AI projects call for a different way of managing work. Agile provides the flexibility and structure needed to handle uncertainty, enabling teams to stay focused and deliver value. Using Agile as your approach to project management for AI projects helps your team handle whatever comes their way while delivering results faster. Feel free to contact Fractal Systems today to discuss how we can help your team adopt Agile for your AI projects.

Want to chat?

We are Fractal Systems Consulting, a product delivery consultancy run by a group of practitioners, professional Scrum trainers, change agents and agile delivery coaches who have deep experience and know-how in creating behavioural change.?

Firms like Schroders, Vontobel Asset Management, Aegon have trusted us to help them achieve their strategic goals faster than ever.

Learn more about how we help global organisations transform their delivery. Register for one of our training courses or contact us at [email protected].