Full Stack Web Developer

Full Stack Web Developer - Application

Contract Software Engineer (AI Applications / High-Velocity Builder)

Location: Remote
Compensation: Contracted hourly
Type: Project-based (20–60 hrs/week)

About Us
We are a stealth AI company building infrastructure that powers reinforcement learning (RL) environments for leading AI labs. Our systems sit directly in the training loop of modern AI — shaping how models are evaluated, improved, and deployed.

This is applied, production-facing work — not experimentation for its own sake.

Role Overview
We’re looking for strong engineers who can own and ship complete applications quickly in fast-moving, low-structure environments.

You should be a high-quality builder first, who also knows how to leverage modern coding tools (including AI assistants / Claude Code / Codex) to increase speed and throughput — without compromising on correctness, maintainability, or system design.

We expect you to use coding agents but not rely on “vibe coding”. You are expected to:

• Understand the systems you build

• Make sound architectural decisions

• Use AI as an accelerator, not a crutch

What You’ll Do

• Build full-stack applications end-to-end (frontend, backend, infra)
• Translate loosely defined ideas into production-ready systems
• Design and implement systems for RL environments and agent workflows
• Ship quickly while maintaining code quality and clarity
• Iterate directly with founders and researchers

How You’ll Work

• One project per engineer - Report to an engineering manager
• You move very fast, but not blindly - All through Slack
• You use tools like Cursor or Claude Code to accelerate development, not replace thinking
• You review, refine, and take ownership of all generated code
• You are comfortable going from 0 → deployed system independently
• You do not need or expect regular guidance from a more senior engineer

This is not a role for someone who wants to ask questions and meet often. This person should be comfortable receiving a project spec document and being able to quickly produce results without the need for oversight.

What We’re Looking For

• Proven ability to build and ship complete applications end-to-end
• Strong generalist engineering skills across:
  • Frontend (React)
  • Backend (Python, FastAPI, Next.js etc.)
  • APIs and integrations
• Solid understanding of:
  • System design and tradeoffs
  • Debugging and performance
  • Writing maintainable, production-quality code
• Experience with:
  • LLMs, agents, or AI-powered systems (preferred)
  • Data pipelines or scraping systems (plus)
• Demonstrated ability to use modern tooling to increase output without lowering standards
• High ownership mindset — operates well with minimal direction

Ideal Candidate Profile

• Has built and deployed multiple real-world products
• Comfortable owning systems end-to-end
• Prior startup, freelance, or rapid-build environment experience
• Uses AI tools pragmatically to move faster — but can build without them
• Thinks in systems, not just tasks

What This Is Not

• Not a narrowly scoped frontend or backend position
• Not a heavily managed or process-driven environment
• Not a project we need to design around the market / user feedback

Why This Role

• Work on infrastructure used by leading AI labs
• High autonomy and real ownership
• Fast iteration cycles and meaningful output
• Remote flexibility
• Opportunity to become a core contributor over time
   
• No long meetings or adaptive to user feedback

Example Work

One example project could be a two-week project where you are asked to rebuild an exact replica of AirBNB

Timeline: 14 days

The app needs to include full functionality of the whole web app, including:

• An interactive map view
• A scrollable search results panel
• Property detail pages
• Realistic mock data seeded into a local database
• Filtering, sorting, grouping, and shareable URL state
• Offline functionality with no external runtime APIs, hosted assets, or external map tile servers

• A production-style frontend/backend architecture

   

Example Tech Stack

• Frontend: TypeScript, React, modern UI library (e.g., Chakra UI), state management (e.g., Jotai)
• Backend: Python (FastAPI), SQLAlchemy
• Database: SQLite (or similar)
• Mapping: MapLibre GL JS (or equivalent)
• Monorepo / Build: Nx (or similar), pnpm
• Linting / Tooling: ESLint, Ruff
• Optional: OpenAPI-generated clients for type-safe frontend/backend integration

The expectation would be that you can study the reference product, identify the core user flows, scope the right features, and build a complete working application end-to-end - All without any oversight or help.

This could include:

• Designing the database schema
• Seeding 1,000+ realistic records with internally consistent mock data
• Building a React/TypeScript frontend
• Building a Python/FastAPI backend
• Implementing map interactions, filters, search state, detail pages, loading states, and empty states
• Matching the visual and functional behavior of the reference application closely
• Ensuring the app can be cloned, bootstrapped, and run locally with minimal setup

The work will be evaluated on scoping, visual fidelity, functional completeness, data quality, and architectural quality.