Developer — Product & Systems

The Role :
Build and maintain the systems that power YFS — and own them end to end. We're hiring a mid-senior developer with a product owner mindset. You won't just execute tickets. You'll receive rough asks from operations, partnerships, and marketing — scope them yourself, build them, ship them, and keep them running. You manage your own queue, communicate proactively, and take responsibility for what you deliver long after launch day. This is a foundational technical seat. The right person sees product, infrastructure, and operations as one job — not three.
 

What You'll Do :
• Ship and maintain the YFS website — new pages, feature releases, performance improvements, and front-end consistency across the full site (Next.js / React)
• Own backend services and APIs — build, deploy, and maintain the server-side infrastructure that powers our programs, mentor matching, and partner integrations, hosted on AWS
• Design and manage databases — schema design, migrations, and query performance across PostgreSQL; working knowledge of AWS RDS, S3, and core infrastructure services
• Build and maintain automation workflows — n8n pipelines covering mentor onboarding, student intake, reporting, and program operations; built to be reliable, not just functional
• Integrate AI into product flows — LLM integrations (Claude, OpenAI) for content generation, feedback automation, and smart matching embedded in real user flows
• Manage your own request queue — receive asks from non-technical colleagues, clarify scope, set timelines, flag blockers early, and deliver without being chased
• Set the technical bar — as YFS grows, review work, define what production-ready means, and support junior developers coming up behind you
 

Who You Are :
• A product owner, not a ticket-taker. You receive a vague ask, turn it into a clear scope, align with stakeholders before building, and own the outcome — not just the output
• Someone who maintains what they build. You've shipped v1 and been accountable through v4. Bug reports, edge cases, and performance regressions after launch are part of your job description, not someone else's problem
• Proactively communicative. Updates go out before they're asked for. Blockers get flagged the day they appear. Non-technical colleagues get plain-English summaries, not jargon
• Comfortable with the full stack. Frontend, backend, cloud infrastructure, database, automation — not a deep specialist in every layer, but competent enough to own features end to end and know when to go deep
• AWS and database literate. You've deployed and maintained real services on AWS (EC2, RDS, Lambda, S3, IAM, CloudWatch) and designed databases that held up under real load — not just tutorial experience
• AI-native. Cursor, Claude, n8n AI nodes, v0 — these are part of your daily workflow. You move faster because of them and you know where to be careful with AI generated code before it hits production
• High agency, low ego. Small team means you'll touch things outside your lane. You'd rather figure it out than wait for someone to assign it

Experience 
• 4–5 years of hands-on product development — we want to see URLs, repos, and systems you can walk us through, not just job titles
• Demonstrated ownership of a product or significant product surface — not just a contributor, but the person accountable for it
• Proven ability to manage inbound requests from non-technical stakeholders: scoping, prioritising, pushing back on scope creep, and delivering on commitments
• Frontend and backend competence — React / Next.js on the front, Node.js or Python on the back; the specific frameworks matter less than your ability to ship across both
• Working knowledge of AWS — EC2, RDS, Lambda, S3, IAM, CloudWatch; you've maintained production services and handled real incidents, not just spun things up
• Solid database fundamentals — PostgreSQL schema design, indexing, migrations, and performance; you understand what's happening below the ORM
• Automation experience — n8n, Make, or equivalent; workflows that handle real data, real failure cases, and real operational dependencies
• AI API integration in production — you've called an LLM API and built something genuinely useful with the output