Frontend and analytics work on a Next.js + TypeScript video AI generator — both the webapp itself and its marketing landing. Backend and model selection were owned by a separate team; my surface was everything users actually touched plus the instrumentation around it.
Stack
- Next
- React
- TypeScript, TS
- Tailwind CSS
- Google Analytics
- Yandex Metrica
- Google Tag Manager
- SEO
- Internationalization (i18n)
Context
A consumer-facing AI video generation product with a webapp and marketing landing. The product needed a polished frontend, granular instrumentation, and the kind of multi-language, theme-aware UI that lets a single codebase serve multiple markets.
Model and pipeline live on the backend side, owned by a separate team. My job: make the surface feel right and make every meaningful action measurable.
What I worked on
- Custom-settings UI for video generation — flexible, with sane defaults but advanced controls available when users want them.
- Theming + multi-language support — single Next.js codebase serving multiple locales without codebase forks.
- Page-level optimisation — perf, layout stability, SEO basics for the marketing landing.
- Analytics layer: Google Analytics + Yandex Metrica + Google Tag Manager, configured so the marketing team can ship new events and tracking changes without engineering tickets.
- "Smart" GTM: events keyed off DOM hooks and dataset attributes — no hardcoded event firing scattered across the codebase, no release needed to add a new funnel step.
Hardest problems
- Making the analytics layer flexible enough that non-engineers could add and adjust events safely — without ending up with brittle tag rules or duplicate firings.
- Multi-language UX without bloating the bundle: locale-aware rendering, correct meta/OG per locale, stable URLs for SEO.
Need a frontend + analytics partner?
I do exactly this kind of engagement: take ownership of the user-facing surface and the instrumentation that proves it works. Reach out via the contacts on my home page.