No-Management Driven Development
Project Management in the AI Era
In a world where traditional workplace hierarchies are being reevaluated, a new paradigm is emerging in the realm of software development: No-Management Driven Development (NMDD). This approach challenges the conventional idea that teams need layers of management to be productive. Instead, it leverages the power of AI to manage decision-making, set priorities, and drive projects forward, all while aligning closely with the preferences of different team members. This model harnesses AI to optimize processes and reduce the reliance on traditional human management structures.
In this blog post, we’ll delve into the core principles of NMDD, exploring how AI can lead to more innovation, faster decision-making, and a significantly more engaged and efficient development team. We’ll begin with a brief review of the evolution of project management and introduce the NMDD approach as a new alternative. We’ll then explore the principles of NMDD and its methodology, present a hypothetical example of NMDD in practice, and conclude with insights into the future implications of this model.
How we got here
Before the dawn of project management, software development was pretty wild. Imagine a bunch of devs diving into projects with nothing but raw skill and a whole lot of caffeine. It was free-form, exciting, but also kind of a mess — deadlines were missed, budgets ballooned, and nobody really knew if the project was on track until it either succeeded wildly or flopped spectacularly. It was clear we needed some kind of structure to keep things from going off the rails.
So, project management entered the scene. It was a game-changer, bringing much-needed order with plans, timelines, and clear roles. But here’s the thing: as software development grew, so did the complexity of managing it. Now, we’re stuck in a world of endless meetings, syncing up more than actually coding, and drowning in so many processes that it feels like we spend more time managing the work than actually doing it.
And let’s not even start on the whole business of estimates. We’ve got tools and techniques out the wazoo trying to predict how long projects will take, but at the end of the day, it’s still a lot of educated guessing. Software development is full of surprises, and no tool can predict them all.
That’s where no-management driven development (NMDD) steps in. It’s like taking a step back to the simpler days of just getting stuff done but with a major upgrade. We’re talking about using AI to fix all those old problems without the heavyweight management frameworks.
As opposed to sprint-based management, which consists of predefined cycles typically spanning two to three weeks and includes structured ceremonies like sprint planning, retrospectives, and effort estimations, NMDD eliminates the concept of fixed cycles altogether. Instead, phases of development, feedback, early detection of potential issues, testing, and seamless deployment start and end constantly, simultaneously, and automatically, adapting on the fly. This seamless workflow is enabled by the power of AI, which fulfills the goals traditionally achieved through sprint ceremonies. From this perspective, the very concept of cycles becomes irrelevant, allowing teams to focus on productivity without the constraints of periodic assessments.
Principles and methodologies
Here are the key principles of NMDD:
1. Autonomy Over Authority (AOA): You’re in the driver’s seat here. Forget about someone breathing down your neck. AI tools manage priorities and guide decisions, tailored to your preferences, enhancing autonomy and flexibility. This empowers you to make informed choices and act swiftly without waiting for layers of approvals. Decision-making becomes streamlined and efficient, as AI provides data-driven insights, helping you focus on what you do best without the burden of bureaucracy.
2. AI as a Collaborative Tool (ACT): Let AI take care of the busywork. Managing schedules, tracking progress, and even taking full responsibility of effort estimations. By analyzing historical data and current performance, AI continuously adjusts its forecasts to ensure that project timelines are always aligned with the team’s actual output, freeing everyone up to focus on the innovative aspects of their work.
3. Continuous Real-Time Feedback (CRTF): Instant updates on how you’re doing, thanks to AI. It’s like having a coach who’s there with you at every step, helping you fine-tune your game in real-time.
4. Adaptive Velocity Management (AVM): In NMDD, velocity is crucial, but it’s managed differently. AI continuously feeds back into the development process, enabling real-time adjustments and improvements. This means effort estimations are constantly updated by AI, removing the traditional obsession with deadlines. The focus shifts from meeting arbitrary dates to maintaining a steady, efficient pace of work. Deadlines become secondary, used only when there are tangible, external constraints that demand a specific timeframe. This approach ensures productivity is evaluated consistently and adaptively, not by sporadic and often inaccurate estimates.
5. Fully Transparent Workflow (FTW): NMDD leverages AI to ensure that every action and decision is visible to all team members. This continuous visibility fosters a culture of trust and collaboration, allowing team members to see the full picture at any time. By eliminating hidden agendas and promoting an open workflow, AI supports a unified team dynamic where everyone can contribute to and benefit from shared insights and data.
A core concept of NMDD is the “AIM Development” methodology, a.k.a “AIMED”. Basically it’s the principal methodology for practicing the NMDD paradigm and its principles. You can think about it as equivalent to “Agile”. It stands for “AI Managed Development”, and it basically aims (pun intended) to shift as much work as possible to AI. It’s a gradual process, and the term is used as an adjective to describe the extent to which an organization or a project automates its processes using AI. You may hear sayings such as: “The development team is highly AIMED”, “This release cycle was more AIMED compared to the previous one”, or “Our project’s workflow is moderately AIMED at this stage” etc.
Putting it into practice
NMDD is a relatively new concept on the tech scene. It’s not yet widely recognized or adopted in the mainstream of software development organizations. As a fresh idea beginning to bubble up in tech circles, NMDD currently lacks the established standards and detailed methodologies that more traditional project management approaches possess. “AIM” is currently quite abstract, and it’ll take some time until it will be associated with specific tools and well-defined processes. It’s very much in the experimental phase, but that’s part of what makes it exciting — the rules are still being written, and the potential for innovation is massive.
Given its immaturity, let’s sketch out what NMDD might look like in practice, even though it’s still being shaped. Imagine the following scenario: a product owner discusses a new feature with a client during a meeting. An AI tool listens in on the conversation, understanding the requirements and nuances of the requested feature. Once the meeting concludes, the AI automatically generates a detailed ticket with all necessary information. This ticket is then assigned to a team member best suited for the task, using AI to analyze past performance, skill relevance, and current workload. For example, if infrastructural work is identified, the AI might generate a separate ticket for the DevOps team.
Once assigned, an AI-based IDE (Integrated Development Environment) plugin on the developer’s (or DevOps etc.) machine recognizes when they start working on the task. It automatically updates the ticket’s status to reflect this. As the developer progresses, the AI continues to monitor work, offering real-time suggestions for improvement, flagging potential issues before they turn into real problems, and automatically creating or updating tests to run in parallel with development. The system ensures that the product owner, team leader, and any other relevant stakeholders are kept in the loop with automated updates, enabling a smooth, continuous flow of information. No status meetings, sync calls, or endless Slack threads; no spending time on writing Jira tickets, and no guessing effort estimations.
This was just a basic example of what NMDD could look like in action. The ultimate goal is to develop an even smarter system — one that not only manages individual tasks but also synchronizes and prioritizes work across all team members involved in a project. This advanced system would ensure that everyone’s contributions align seamlessly, optimizing the overall workflow and dramatically increasing productivity.
Fortunately, we’re not starting from scratch. There are already numerous AI tools available that can be leveraged to build such a system. These tools range from AI-enhanced project management software to smart IDE plugins and automated update systems, all designed to integrate smoothly and provide a robust foundation for an NMDD environment. As these tools continue to evolve and become more sophisticated, the path to a fully realized NMDD system becomes clearer, promising a revolution in how software development is managed and executed.
Conclusion
NMDD, or no-management driven development, is about cutting back to the essentials of what made coding cool in the first place, but with a twist: leveraging AI to keep the chaos at bay. It’s streamlined, it’s efficient, and it lets us get back to what we do best — building awesome software.
