As an Engineering Manager, I’ve long struggled with the mental overhead of sprint planning across multiple specialty teams—capacity lives in my head, adjustments for holidays are manual, and it all bottlenecks on EM experience. In this post I share a lightweight script that calls the OpenAI API (with some prompt engineering) to:
Analyze the past 2–3 sprints’ velocity per specialty
Auto-adjust for bank holidays
Recommend a realistic story-point capacity range
The result: democratized planning, reduced bias, and empowered engineers to run their own sprints. Would love to hear your feedback or collaborate on pushing agile planning further with AI!
In my journey as a software engineer manager, I've witnessed the complexities of engineering resource planning firsthand. The challenges are multifaceted: managers must account for bank holidays, employee absences, holiday allowances, and sick leaves. Additionally, some team members split their time between management and individual contributions, further complicating the equation.
