In chapter 8 of Pascal Bornet’s book Agentic Artificial Intelligence he discusses how to be successful in your implementation of AI agents. Agentic AI promises to transform how organizations operate, but not every opportunity is worth pursuing. Some projects are too costly, too complex, or simply not feasible today. In his book Agentic AI, Pascal Bornet offers a simple but powerful framework for identifying where to focus first: the Three Circles of Agentic Opportunities.

This model helps us spot the “sweet spot” — the opportunities that are not just visionary, but practical, impactful, and achievable. We have previously looked at what Agentic AI is. We’ve considered the foundational keystones tyo help us understand what AI agents are and now we are ready to look into our organization adn look for opportunities to actually begin using AI agents.

Hight Impact

Does this opportunity deliver meaningful value? Why does it matter? Impact could mean increased revenue, reduced costs, improved customer experience, or broader social benefit. The higher the impact, the more transformative the outcome. Look for routine tasks that are preventing your best people from doing their best work. When assessing the impact, you could actually use a point system of 1 to 5 points to rate time saved, value added, error reduction and scalability impact (can we do this across the organization?).

High Feasibility

Can it actually be done today? High feasibility means the data is available, the technology exists, and the skills are within reach. Chasing moonshots is tempting, but feasibility keeps teams grounded in what’s realistically possible. Do do have clear consistent rules for completing a task? Is the input data and information accessible? If something goes wrong, is it easily managed? Are the results of the task easily observed and measured? You want to be able to explain the rules without saying too many “it depends”. Data analytics projects that consume data and produce reports may be good candidates. We can use a point system to rate this as well.

Low Effort (Low Cost)

How much work will it take to get results? Is it worth it? Low-effort opportunities require minimal investment, shorter development cycles, and fewer organizational hurdles. They often bring faster wins that build momentum.Is the process well-documented? If it seems like it’s too much effort, can you find a way to start small and scale up later?

The Sweet Spot

Where these three circles overlap is the sweet spot of agentic opportunities. Projects in this zone deliver: For example, automating meeting notes and follow-ups may not sound revolutionary, but it’s high impact (saves hours of knowledge worker time), low effort (leverages existing tools), and highly feasible (mature technology already exists). Contrast that with developing a fully autonomous research scientist agent — high potential impact, but also extremely high effort and currently low feasibility for most organizations.

A Methodology

Pascal Bornet identifies four steps to identifying your agentic opportunites. Task inventory, impact assessment, feasibility assessment, and implementation effort.

Why This Resonates with Business Analysis

Bornet’s steps may feel familiar — and they should. In fact, they echo long-established practices in business analysis, such as those found in the BABOK® Guide. BABOK goes into far greater detail, but Bornet adapts the spirit of those methods to the new frontier of agentic AI. That’s an important reminder: agentic AI is not starting from scratch. It builds on decades of proven business analysis and process improvement thinking, while adding the new lens of intelligent agents.

Conclusion

Finding the right agentic opportunities is about more than chasing shiny technology — it’s about aligning impact, effort, and feasibility. With the three circles for clarity, and the four steps for action, leaders can confidently identify where to start. Every organization needs to make a business case for a new project. That almost always involves a cost–benefit analysis:

• Cost is reflected in effort and feasibility.
• Benefit is reflected in impact.

Pascal Bornet’s framework helps translate these timeless business principles into the emerging world of agentic AI. By using cost–benefit thinking as the anchor, leaders can ensure their first agentic AI projects are not only exciting, but also credible, fundable, and sustainable.

Next Steps

After you have identified the right AI agent opportunuities, we need to define what type of agent we need. Pascal Bornet defines levels of AI agents. Often we may find that we are choosing between a Level 2 or Level 3 agent.

More Information

Are you looking for a bit more guidance on finding agentic opporotunities? Finding Agentic AI Opportunities by Industry.