When artificial intelligence started making headlines in project management circles, many of us were intrigued — and a bit skeptical. Could an algorithm truly support the nuanced, often chaotic nature of managing people, timelines, and scope? In this hands-on article, I explore how AI is not replacing but augmenting my work. Drawing from direct experience, I highlight how project teams — from directors to developers — can use AI to boost clarity, speed, and strategic value in every phase of project delivery.
Over the past year, I had the opportunity to participate in the initial phase of experimenting with AI tools in real projects within my company. My goal was to explore how AI could support both my productivity and the quality of the team’s work. I focused on practical integrations — tools that could solve specific problems we face every day. Most importantly, I treated every AI generated result as a draft — a suggestion, never a final answer. My interest in AI started from a simple need: I was tired of spending hours on repetitive documentation and reporting. I wanted to focus more on thinking, leading, and solving real problems — not formatting slides or writing status updates.
Practical use cases
To better understand AI’s real impact on everyday project work, I focused on three key areas where it could bring visible, measurable value. These practical use cases span communication, documentation, and delivery — and they’ve reshaped how I lead projects today.
Meeting Notes in Minutes, Not Hours
Back-to-back meetings are part of the routine — and so is the burden of turning them into structured summaries. I used to spend 75 to 90 minutes crafting meeting minutes with clear action items and responsibilities. Now, using Microsoft Teams transcripts together with SafeBrain and Microsoft Copilot Standard, I generate draft summaries in just a few minutes. Copilot organizes the content, while SafeBrain extract decisions and action points. AI can generate notes in any language, as long as the right settings are applied. It can even provide adapted translations, improving clarity across multilingual teams. That said, I never share these summaries without a full review. Sometimes AI misses context, mislabels a speaker, or oversimplifies decisions. My role is to validate and refine — turning a fast draft into a trustworthy final version. What I’ve learned: with clear input and careful review, AI bridges communication gaps and saves time — but it’s still on us to ensure the content is accurate.
First Drafts, Better Questions, and Smarter Acceptance Criteria
Clear documentation drives collaboration — and the earlier in the process, the better. To improve the quality of sprint planning and user stories, I began using Microsoft Copilot and built-in tools like Atlassian Intelligence to generate initial drafts of reports and structured updates. For example, I often input raw Jira data, and the AI suggests themes and highlights progress and identifies blockers. But where AI truly made an impact was in crafting user stories. Using SafeBrain, we generated clarification questions for the Product Owner, ensuring we fully understood the scope before writing acceptance criteria. The team then reviewed and finalized the stories collaboratively. What I’ve learned: AI gives us momentum, but clarity still comes from thoughtful team discussion. Good documentation begins with better prompts — and ends with team validation.
Estimation with AI: When Jira Talks Back
Once we improved story quality, we turned to effort estimation. Using AI-enabled Jira plugins like Atlassian Intelligence, we cross-checked our sprint plans against historical patterns. In one case, the AI flagged our 12-week delivery estimate as too optimistic. That prompted the development team to propose their own AI experiments to help increase efficiency. I supported their initiative and secured budget approval from senior management. They tested coding assistants that helped increase both productivity and delivery quality. Still, no AI outputs were used blindly. Everything — from the estimation corrections to code, tests and rewritten stories — was reviewed by the team. AI inspired better decisions, but never replaced them. What I’ve learned: when roles like Project Directors, Project Managers, Product Owners, and Developers adopt AI collaboratively and critically, the entire delivery chain benefits — without compromising accountability.
AI Adoption Across the Team
Of course, no experimentation happens in isolation. As we began testing AI across different project layers, the team itself became part of the journey — adapting, responding, and ultimately shaping how these tools were used day to day. With these early wins, interest in AI grew across the team — though not without hesitation. Some were skeptical, unsure whether these tools would help or distract. What worked best was leading by example and sharing real gains, such as a 70% drop-in time spent writing meeting notes. We also encouraged open conversations about bias, errors, and human review. That created a safer space for the team to explore AI as a support tool — not a decision-maker.
From Experiment to Scaled Practice
The AI experimentation phase lasted around two to three months. During this time, we explored real use cases, refined our approach, and built the skills and confidence needed to move forward. We also measured impact through concrete indicators — such as time saved on documentation, reduction in planning errors, and faster decision turnaround. Once the value became evident, I used the experience, tools, and lessons from that pilot period to implement AI practices across all the projects I currently lead. That foundation — built together through curiosity, trial, and reflection — made the transition to full-scale adoption both natural and effective.
Lessons Learned: Tips for Starting with AI
These experiences have taught me that success with AI is less about the technology and more about how you introduce it to the team.
Start small, but think system-wide
Introducing AI through a single, well-chosen use case — like meeting notes or estimation — helped build confidence without overwhelming the team. Once the value was proven, it became easier to scale adoption across planning, communication, and reporting. Small wins created momentum for broader adoption.
Treat AI outputs like drafts, not directives
No matter how accurate the suggestion seemed, we never acted on AI-generated outputs without review. In one instance, an automated project summary missed key stakeholder concerns — something only human context could catch. This reinforced the principle that AI can accelerate work, but people are still the ones accountable for its quality and implications.
Use AI to improve clarity
One of the most underestimated benefits of AI was its ability to surface gaps in our thinking. Tools like SafeBrain helped us ask better questions during story refinement, which led to sharper acceptance criteria and fewer misunderstandings during development. In this sense, AI became less a shortcut — and more a mirror.
Make AI a team habit
The turning point wasn’t when I started using AI — it was when the whole team began to see it as a helpful companion rather than a threat. We created shared workflows, openly discussed what worked or failed, and normalized human oversight. That collective approach made adoption smoother, more consistent, and ultimately more sustainable.
Final Thought
While these experiments are still ongoing, one thing is clear: AI has already changed how I work, how my team collaborates, and how I think about leadership. AI didn’t replace me as a project director — it made me better. But only because I approached it with curiosity, accountability, and collaboration. Every story, every estimate, every communication still reflects human insight. The value of AI isn’t in automation alone. It lies in augmentation — empowering teams to move smarter, faster, and with greater clarity, while staying grounded in real human leadership.
Project Director and Agile Coach with 22 years of experience leading complex digital and agile transformations. PMP®, PMI-ACP®, ASPC certified and active mentor in global PM and Agile communities.