The challenge
Adding a coding assistant is straightforward. Introducing a useful development practice across a team takes more work. Requirements, architecture, implementation, review, and testing still need to fit together, and people remain accountable for the result.
At Applied Visions, my work included helping introduce AI-assisted delivery practices across approximately 30 engineers and more than 20 projects.
My role
I worked across technical strategy, team adoption, and practical implementation. The scope included requirements intake, story refinement, estimation, architecture planning, coding assistance, review, testing, documentation, and production support.
I also advised client organizations on agentic development through workshops, playbooks, pilot projects, and strategy conversations.
The approach
The work treated AI adoption as part of the delivery process. Generating an implementation is one step. Understanding the requirement, inspecting the change, testing its behavior, and deciding whether to ship remain engineering responsibilities.
The emphasis was on making the practices usable across actual projects and keeping technical ownership clear.
Outcomes and boundaries
The adoption scope is documented in my professional background. This overview deliberately avoids a blanket productivity percentage. Measuring delivery speed, quality, and rework requires project-specific evidence and a consistent baseline.
The practical question remains: does this workflow help a team deliver software it understands and can maintain?
Selected outcomes
- Delivery practices introduced across approximately 30 engineers and more than 20 projects.
- Adoption extended beyond coding into planning, review, testing, documentation, and support.