Easycoder worked with AWS to deliver a two-day workshop that helped MILK Books turn real development priorities into practical solutions and hands-on experience with AI-DLC.
For engineering teams, the opportunity with AI extends well beyond generating code. The bigger question is how to connect AI’s capabilities with business requirements, existing systems and the decisions that shape good software.
That was the focus of Easycoder’s recent engagement with MILK Books. Working with the AWS team, we delivered a hands-on workshop using AWS’s AI-Driven Development Lifecycle, or AI-DLC. The aim was to apply the methodology to real work and help the customer’s engineers develop experience they could carry into future projects.
A structured approach to AI-driven development
AI-DLC positions AI as an active collaborator throughout software delivery. It can break down a business goal, ask clarifying questions, propose designs and generate implementation work. People retain responsibility for reviewing outputs, resolving trade-offs and approving critical decisions.
AWS defines three connected phases:
Inception: turn business intent into requirements, acceptance criteria and manageable units of work. Teams review and refine AI’s proposals together.
Construction: develop those units through design, code and testing, with engineers validating the outputs.
Operations: support deployment, monitoring and maintenance through AI-generated recommendations and human oversight.
Requirements, plans and design decisions provide persistent context between stages. The method also uses short development cycles called “bolts”, with frequent validation and feedback. For existing applications, understanding the current codebase is an important part of the process.
This creates a framework for applying AI across delivery while keeping the team involved in the decisions that matter. AWS’s explanation emphasises both AI-powered execution and real-time team collaboration, making shared problem-solving central to the approach.
Applying the method to real customer priorities
The MILK Books workshop brought two engineering teams together to tackle existing backlog priorities. The work centred on improvements to the customer’s digital platform, giving participants an opportunity to explore AI-DLC within a codebase they already knew.
Over two days, the teams progressed from defining the challenges to developing practical solutions, applying AI-DLC to real priorities in their existing codebase. Business framing remained part of the development conversation, helping connect the technical work with the purpose behind it.
One useful observation was AI’s ability to map the existing codebase and connect information across it. Using the current implementation to inform requirements helped the teams work with the system’s existing behaviour and constraints.
The collaborative format also delivered value. Bringing people together surfaced ideas and shared learning that would have been harder to develop through individual experimentation.
Practical experience to build on
By the end of the second day, both teams were productive with the AI-DLC method and intended to use it in future projects, adapting the workflow to suit how they build.
The workshop also highlighted what supports effective adoption. Clear problem statements, strong documentation and accessible context helped give the AI better direction. Keeping plans and progress visible, and maintaining team ownership of the solution, emerged as important considerations for future use.
The teams gained practical experience of how to guide AI, evaluate its proposals and collaborate around a shared development goal.
For Easycoder, the engagement reinforced the value of starting with a specific customer priority and learning through delivery. Working with AWS and the MILK Books team created a practical setting to explore how AI-DLC can fit into everyday engineering work.
Have a development challenge or an AI project in mind? Let’s talk.
