BetterFrontend.dev

Problem

Most people learning frontend development are stuck choosing between passive tutorials that don't transfer to real work, or expensive bootcamps with generic curricula. What's missing is the thing that actually builds skill: working on a real, self-directed project with an experienced developer reviewing your work and unblocking you along the way. BetterFrontend.dev set out to be that, a platform where learners bring their own project ideas and get paired with a mentor for hands-on, structured guidance instead of a fixed syllabus.

My role

I founded BetterFrontend.dev and led development, from the initial architecture through the mentor-student experience itself. I collaborated with three senior engineers, one junior engineer, and one designer on feature work, while owning the mentorship model: how mentors and students get paired, how project reviews happen, and how progress gets tracked between them.

Approach

Built on a T3-stack foundation, using Next.js App Router, tRPC, and Prisma over PostgreSQL for a type-safe path from the database to the UI. Sign-up ran through GitHub OAuth via NextAuth, which doubled as a soft filter for developers already comfortable with the tooling they'd be using.

Paid mentorship tiers were metered through Stripe subscriptions. PostHog tracked product analytics and login/registration funnels from day one, Sentry covered error monitoring, and Playwright ran end-to-end coverage on the critical paths (registration, onboarding, project creation). The product shipped in English, Turkish, and Ukrainian via next-intl, to reach developer communities directly instead of only in English.

Outcome

Ran as a public beta from January 2025 through August 2025. Iterated on the pricing and free-tier model more than once based on early user feedback, and shipped the mentor-student project workflow described above. Registration has since been closed and the project is being wound down as the domain lapses.