Prototype Engineer

What You'll Do

• Design and develop tools that expose AI/ML capabilities to autonomous agents, building the orchestration layer that solves complex domain use-cases in production homebuilding.
• Take proven research outputs and integrate them into the Higharc product as packages and components, following best practices established in a large, complex codebase.
• Work across the Python ML/AI service layer and the TypeScript/React product, wiring new capabilities end-to-end via API.
• Build frameworks for rapid evaluation, iteration, and promotion of prototypes.
• Operate in lock-step with the broader engineering organization and participate in customer-facing labs sessions, using direct customer signal to validate prototypes and inform iteration priorities.
• Serve as the primary translation layer between the research team and the product engineering team. Convert experimental notebooks, fine-tuned models, and RAG pipelines into maintainable, testable, deployable code.
• Develop working knowledge of the homebuilding domain sufficient to make independent judgment calls about which capabilities matter, which edge-cases are load-bearing, and where AI falls short in the built environment.

About You

Required:

• 4+ years of professional Software Engineering experience, with 2+ years integrating ML/AI capabilities into production products
• Shipped AI-powered features to real users, not just model training
• Experience working in or adjacent to a research team
• Professional proficiency with Python, Typescript/React, API design, LLM tooling, evaluation design, and AI calibration
• Familiarity with LLM APIs (OpenAI, Anthropic, or open-source), vector databases, RAG architectures, Python ML ecosystem (PyTorch, HuggingFace), modern frontend frameworks (React/Next.js), CI/CD and testing frameworks

Major plus:

• Experience in construction, AEC, or other physical-world domains with messy, under-structured data
• Experience with agentic frameworks (LangChain, CrewAI, or custom harnesses)