About the Role:

We are building the next generation of software creation, a fully autonomous AI Coding Agent that can plan, code, debug, and deploy full-stack applications from natural language.

We are looking for a Founding Engineer who understands how code is parsed, transformed, executed, and validated. You will build the core engine that turns prompts into running, production-grade apps and set the technical foundations the entire product will be built on.

Note: You don’t need a formal compiler background. Strong systems intuition and real-world tooling experience matter more than academic credentials.

Requirements:

• Deep Systems Understanding: You possess strong knowledge of Node.js/runtime fundamentals, file systems, module resolution, build tools, and how code is parsed/executed.
• AST & Tooling Experience: Hands-on experience with AST tooling (Babel, SWC, Acorn, Tree-sitter) or similar programmatic code transformation systems. You understand that reliable code editing requires more than regex.
• AI Engineering: Experience building agents using tool/function calling and structured outputs (JSON schemas). You have built systems with real feedback loops, not just simple prompt chaining.
• Builder Mentality: You have shipped developer tools (CLI, plugins, internal platforms) or meaningful open-source work and enjoy the chaos and creativity of building from zero to one.

Responsibilities (What You Will Build):

• Execution Sandbox & Preview Pipeline: Design a secure, scalable code execution workflow using Vercel (serverless/edge), handling isolation, resource limits, and safe dependency management.
• AST & Code Manipulation: Build systems that "understand" code using ASTs to enable safe edits, refactors, lint fixes, and structured transformations.
• Real-Time Feedback Loop: Implement type-checking, linting, and automated test feedback so the AI agent can detect issues and self-correct before the user sees them.
• LLM Orchestration (Agentic Loop): Own context management and tool calling. Implement RAG over a dynamic file system and determine when the agent should plan, write, or debug.
• Technical Ownership: Define architectural patterns, establish best practices, and guide future engineers as the core team grows.

What Success Looks Like (First 90 Days)

• A reliable execution + preview pipeline running real user-generated apps.
• The AI agent can safely modify multi-file projects using AST-based transformations.
• A clear agentic loop with measurable improvements in success rate and iteration speed.
• Foundational technical decisions documented and ready to scale.

Nice to Have (Bonus Points):

• Experience with CRDTs (Yjs / Automerge) for real-time collaboration.
• Rust experience for high-performance tooling or parsers.
• LSP (Language Server Protocol) experience.

What We Offer:

• Remote Culture: Work from anywhere.
• High Impact: You are building the engine, not just a feature.
• Modern Stack: Work with the latest in AI, Serverless, and AST tooling.
• Competitive Equity & Salary: We are building a unicorn, and we want you to own a piece of it.