Summary
Brooks' Law suggests that adding more people to a late software project can make it even later. But what happens when we replace people with AI agents? Does adding more agents always improve productivity?
As AI adoption grows, organizations are introducing agents for development, testing, reporting, planning, and project management. However, more automation can also introduce new challenges, including coordination overhead, duplicated work, and additional validation.
Drawing on my experience managing distributed Drupal teams and exploring AI-powered delivery workflows, this session challenges the assumption that more AI always means faster delivery.
Problem Statement
While managing enterprise Drupal projects, I've experienced situations where adding more resources didn't necessarily speed up delivery. Dependencies, unclear ownership, communication gaps, and repeated reviews often created additional challenges.
As I started exploring AI agents for project management, an important question emerged: Could adding more AI agents create similar coordination problems?
Without a clear strategy, teams may spend more time reviewing and managing AI-generated outputs than the automation actually saves.
Approach / Benefits
Using Brooks' Law as a starting point, I'll explore how teams can introduce AI agents without creating unnecessary complexity.
Through practical Drupal delivery examples, I'll compare smaller and larger AI-assisted workflows and examine how coordination, dependencies, accuracy, and human review affect overall efficiency.
I'll also introduce a simple Agent Coordination Equation to help evaluate whether an additional agent delivers real value or creates more work.
The goal is to help teams use the right agents for the right activities, rather than introducing more agents simply because they can.
Learning Objectives
- How Brooks' Law can help us rethink the adoption of AI agents.
- Why more AI agents don't always lead to faster delivery.
- How to identify coordination challenges and dependencies between agents.
- How to evaluate time saved against validation and rework effort.
- How to decide when adding another AI agent is beneficial.
- How to balance automation, efficiency, and human oversight.
Why Should People Attend?
More AI doesn't always mean better results. This session will help you understand the hidden challenges of AI automation and design workflows that improve productivity without creating new bottlenecks.
Target Audience
- Project / Program / Delivery Managers
- Product Owners / Scrum Masters
- Engineering Managers / Technical Leads
- Agency Leaders / Executives
- AI Solution Architects
- Business Analysts / Automation Enthusiasts