My role in practice

I worked as a frontend team lead, helping the team turn proposed work into something that could be planned, reviewed and delivered reliably.

At the beginning of the engagement I reviewed AI-generated stories, worked through timelines and dependencies, and helped identify where delivery relied on other teams or external work.

I also worked closely with the client team throughout delivery, including regular catch-ups with my client counterpart to discuss progress, expectations, areas that needed attention and ways of improving the development process.

Planning and dependency awareness

The programme involved multiple teams and moving parts, so delivery was not simply a matter of implementing isolated tickets.

Part of my role was to understand dependencies early, question whether proposed work was feasible within the available time, and make sure the pod had enough context to make sensible implementation decisions.

Raising the quality bar

A significant part of my work centred on merge request review and delivery quality.

I reviewed merge requests for TypeScript validation errors, build failures, implementation quality and, where appropriate, accessibility issues. For visual or behavioural changes I would often pull branches locally so I could review the actual component, layout and interaction rather than relying only on the code diff.

This added scrutiny to individual reviews, but it also reduced the number of avoidable issues reaching the client review stage.

Over time, recurring TypeScript, build and quality problems became less common as the team became more familiar with the expected process and standards.

Scaling across locales

The checkout needed to support different behaviour across international markets without allowing locale-specific rules to turn into a growing collection of conditionals.

We used a Strategy-pattern approach to separate market-specific behaviour, helping keep the implementation more maintainable as functionality varied between countries.

Working across team and client boundaries

The role also involved acting as a bridge between agency and client teams.

When expectations or implementation approaches differed, I helped surface the issue early, clarify what was needed and keep discussions focused on practical delivery rather than allowing friction to accumulate around review.

As quality improved and issues were addressed earlier, the working relationship became easier and client confidence in the delivery process improved.

AI-assisted delivery with human judgement

AI tools, including OpenAI Codex and ChatGPT, were embedded heavily in the development workflow and were important in helping the team work at the pace required.

The engagement also reinforced that faster code generation does not remove the need for senior judgement. Sequencing work, understanding dependencies, reviewing implementation quality, protecting accessibility and aligning expectations still required people who understood the wider system and delivery context.

Outcome

The review process helped improve consistency and reduce recurring build, TypeScript and implementation issues across the team.

The work also helped create a clearer quality bar for delivery while supporting a large, multi-team checkout modernisation programme operating across international markets.