Founding Engineer (Full Stack)

We need someone with 2+ years of full-stack engineering experience who has built something real from 0 to 1. You should be comfortable working in a fast-paced, lab-like environment and have a track record of solving complex algorithmic problems. Bonus points if you have computer vision/ML experience or previous startup/founding experience. What you'll do: - Build and train custom computer vision models and fine-tune LLMs for construction automation - Take ownership of the intelligence layer and drive algorithmic solutions rather than just application plumbing - Work closely with the founding team to orchestrate multiple AI models into cohesive systems - Architect full-stack solutions using Python, React/TypeScript, and our unique tech stack (Prefect, Supabase) - Lead technical projects from conception to deployment that directly impact construction industry workflows - Design and implement complex systems that require creative problem-solving beyond standard algorithms. Day-to-day example: You might start your week debugging a computer vision model that processes construction blueprints, work with the team to orchestrate 7 different AI models into a unified workflow, and spend time designing algorithms to automate tedious manual processes. Last month, our team tackled automating blueprint analysis that typically takes hours of manual work, and the person in this role would have built the core ML pipeline and full-stack interface to make it production-ready. This is the kind of challenging, intelligence-focused work you'd thrive in. Our Tech Stack: - Backend: Python, FastAPI, Prefect (workflow orchestrator) - Frontend: React, TypeScript - Infrastructure: Supabase (database + edge functions) - AI/ML: Custom computer vision models, LLM fine-tuning, multi-model orchestration - Development: Heavy use of Claude, Cursor, and other AI coding tools. What Success Looks Like: - You've built a product that people actually used (10K+ users or paying customers) - You can solve our math-heavy technical interview better than our founder - You can quickly onboard to our unique tech stack and contribute meaningfully - You're excited about doing the "fun part" of engineering - algorithms and intelligence rather than just plumbing. Interview Process: 1. 15-minute phone screen 2. 45-minute technical screen (open-ended problem solving, not leetcode) 3. ~20-hour take-home project (full-stack problem that mimics our real work) -- can be live technical instead as well 4. Multi-day onsite (flexible on length based on signal strength)