Problem
Repetitive work around the code (creating tasks, opening and describing pull requests, code review, taking new features from plan to release) took up the team’s time and was done differently by different people.
My role
Creating and rolling out AI skills and automations for the development team that standardise this work and cut the time it takes.
What I built
- Automated creation of Jira tasks from business documentation.
- Opening pull requests together with a description that follows the company’s convention.
- Code review support: a project knowledge base the agent relies on, plus dedicated code review skills.
- AI SDLC: the whole feature lifecycle supported by AI, from planning to implementation, based on Matt Pocock’s AI Hero process (aihero.dev) and adapted to the way the team works.
- MCP servers used in many parts of the process, so the agents work directly with the team’s tools.
Outcomes
- Repetitive work around the code gets done faster and to one consistent standard, whoever does it.
- One shared, AI-supported process for taking features from plan to implementation.
- Tools used by the whole team, not just by me.