AI Engineer (Part-time)

Part-time AI Engineer for mobile (Flutter) and web products. 20 hrs/week, fully remote. Use AI tools to solve product tasks end-to-end.

We build mobile apps for iOS and Android, plus web products, monetized via subscriptions and ads. About half of projects are Flutter mobile apps, half are web. We need someone who can quickly dive into a product and close tasks using AI. What to do: - Take a task on any of our products (mobile or web) and bring it to release - Build simple Flutter apps for iOS and Android: from spec to App Store and Google Play publication - Build and maintain web parts of projects - Implement layouts from Figma - Integrate subscriptions and paywalls (AppHud), analytics and attribution (AppsFlyer, Firebase), ads (AdMob) with mediation - Work with remote config, feature flags, A/B tests

AI tools: Claude Code, Cursor, Codex — at least one in daily use (details below). Holistic product understanding: how client communicates with server, APIs and authentication, where data lives, what happens from button tap to DB write. Can diagnose which layer broke and inspect logs and network requests. Working with unfamiliar codebase without long onboarding. Flutter: built apps, understand internals, experienced with flutter_bloc. Build and releases: signing, profiles, keystore, App Store Connect, Google Play Console, both platforms. Web: any modern stack; willingness to work with existing code matters more than specific framework. Layout: convert Figma designs into screens that work across screen sizes. MacBook (required for iOS builds). AppHud, AppsFlyer, AdMob experience is a plus — but willing to learn is fine too. About AI — read carefully: We use AI extensively. In your application, describe in detail how you use AI in work and general life. Applications without this part are not read. Interesting: what tools and for what; workflow (instructions for agents like CLAUDE.md, custom commands, skills, hooks, MCP servers, parallel sessions); first steps in unfamiliar project; real tasks solved via AI and failures; how you verify generated output and catch hallucinations before release; AI beyond coding.

20 hours per week, 4 hours each working day. Available during working hours. Fully remote. Compensation determined after interview based on experience. Process: HR screening → test assignment → call with team lead.