Full Stack Engineer
Engagement: Freelance / contract only
Workload: Depends on the project — typically 2–6 months per engagement, often with follow-on work
About the role
A growing part of our pipeline isn't websites — it's products. Clients come to us with a working prototype (often AI-generated), a locked UX and a launch date, and need someone to take it to production: web apps, mobile apps, multi-tenant platforms, admin panels, AI-assisted features.
You'll be the engineering counterpart to our design team on these projects — scoping, architecting, building and shipping.
What you'll do
- Take products from prototype/MVP to production: two-sided platforms, patient and provider portals, marketplaces, membership platforms
- Build backends: REST/GraphQL APIs, auth and roles, multi-tenancy, payments and subscriptions, file and document storage, notifications
- Build admin and CRM panels: user management, analytics, GDPR data deletion, crash reporting, invoicing
- Integrate third-party systems: hardware and device APIs and webhooks, payment providers, CRMs, analytics warehouses (Snowflake, Tableau)
- Build AI features into products: LLM APIs, assistants scoped to product data, orchestration, guardrails
- Debug and harden inherited codebases — several clients arrive with API and webhook bugs from previous contractors or vibe-coded prototypes
- Ship cross-platform mobile apps and handle App Store and Google Play submission
Scope work and produce hours-based estimates with a breakdown by phase and risk
Must-have
- 5+ years full-stack, with production ownership rather than just feature work
- Backend: Node.js/TypeScript or Python — APIs, auth, background jobs, webhooks
- Frontend: React (Next.js a plus)
- Databases: relational schema design, PostgreSQL or similar. Supabase or Firebase welcome
- Payments and subscriptions in production (Stripe or equivalent)
- Deployment and infra basics: CI/CD, environments, monitoring, logging
- At least one product taken from zero to live users
- AI as a core part of your toolkit — in two directions. First, how you build: Claude Code (or equivalent agentic tooling) for scaffolding services, refactoring, test generation, and rapidly assembling working prototypes to pressure-test an architecture before committing to it. Second, what you build: hands-on experience integrating LLM APIs into a product — prompt design, structured output, tool use, evaluation, cost and latency control. You should also be able to audit an AI-generated codebase and honestly assess what's salvageable, since a growing share of our clients arrive with exactly that.
- English B2 or higher — you'll present the technical approach on client calls
Ability to break a brief into a phased estimate with hours and risks
Nice to have
- Cross-platform mobile: React Native or Flutter, plus store submission experience
- Compliance-adjacent work: HIPAA, GDPR, SOC2 prep — we have healthtech and fintech leads
- Real-time features: WebSockets, LiveKit, streaming
- Data and analytics integrations: Snowflake, Tableau, event pipelines
Web3: smart contract integration, wallet auth
Soft requirements
- You give honest estimates and flag when a client's budget doesn't match the scope — plenty of our leads arrive with a 250-hour budget for a 500-hour product
- Comfortable talking to non-technical founders and translating business logic into architecture
You can run technical discovery: ask the right questions, produce a scoped plan
Application
Send 2 products you took to production with your role and the stack, an example of AI tooling you've used in your build process or shipped into a product, your rate and preferred engagement model.