Founding Head of Engineering (0→1 / Vertical AI / Construction Tech) - Stealth AI
11 days ago
Boston
Job Description Overview We're building a new vertical AI company serving the construction industry. Our first products will use LLMs, agentic workflows, and computer vision to automate painful, manual workflows involving project procurement documents and construction submittal packages—catching errors before they become expensive mistakes in the field. We're hiring a Founding Head of Engineering to architect and ship the first version of our platform, partner closely with founders on product direction, and build the foundational engineering culture and team from the ground up. This role is the right fit for those who thrive in ambiguity, love turning messy real-world workflows into software, and want to own an entire technical function from greenfield to scale. What you'll own • Greenfield system architecture & product build, • Decide the early stack and technical direction, • Lead design and development of backend systems, data pipelines, and application logic, • Build the initial user-facing product along with the team you create., • AI + Agentic systems, • Design systems that evaluate and integrate LLMs, document parsers, embeddings, retrieval, and chain/agent orchestration, • Build pipelines that take in messy PDFs, drawings, and submittals and turn them into structured, actionable intelligence, • Make pragmatic decisions on when to build, when to buy, and when to leverage foundational AI models, • Cloud & platform buildout, • Own the creation of cloud infrastructure from scratch: environments, CI/CD, monitoring, authentication/authorization, data storage, security, and deployment automation, • Balance shipping velocity with reliability and future scalability, • Technical leadership & team-building, • Be the principal IC early—hands on the keyboard daily, • Build the engineering culture, set standards, enforce quality, design review processes, and project tracking rhythm, • Recruit, hire and mentor early engineers (founding ICs, ML/LLM engineers, and frontend/backends) with the assistance of a technical sourcer., • Serve as technical partner to the founder and head of product, shaping roadmap and tradeoffs, • Execution in uncertainty, • Own delivering a V1 product that solves real problems—even if the product direction evolves rapidly, • Has taken a product from a blank slate to a production-ready system, including making foundational technical decisions (architecture, stack, infrastructure)., • Experience owning early-stage development end-to-end—writing initial code, deploying to cloud infrastructure, and supporting first customers., • Built in the last 12-18 months (when modern LLMs became practical), • Shipped RAG systems, document parsing with LLMs, or agentic workflows, • Can discuss prompt engineering, evaluation, and model selection tradeoffs, • Design a database schema for complex domains, • Build scalable APIs (REST, GraphQL, or similar), • Set up async job processing and queues, • Built OCR pipelines or document parsing systems, • Worked with PDF extraction, table detection, layout analysis, • Experience at document intelligence companies (Instabase, Docugami, Rossum), • Built computer vision systems for image classification, object detection, or segmentation, • Experience building multimodal AI applications, • Worked with construction-related CV applications: blueprint analysis, site photo processing, safety compliance detection, progress tracking, • Experience with vision models (YOLO, SAM, CNNs) or multimodal LLMs (GPT-4V, Claude with vision), • Built ML pipelines for training, evaluation, and deployment of custom models, • Experience with spatial data processing, 3D reconstruction, or point cloud analysis, • Worked at Procore, Autodesk (construction), BuildingConnected, PlanGrid, OpenSpace, Fieldwire, or similar, • Built products for contractors, subcontractors, or construction management, • Understands construction workflows, terminology, and pain pointsWho you are, • Energized by building from scratch and shipping fast, • Thrives in undefined spaces and helps turn chaos into clarity, • Deeply technical, but also pragmatic—you value use cases and outcomes over academic elegance, • Comfortable being the expert in the room and the person writing the code, • Collaborative and candid—strong communicator with product, customers, and founders Every procurement mistake and every submittal mismatch costs real money and delays. We're building AI systems that prevent those failures before they happen. If you want to: • Own the technical build from day one, • Build a world-class engineering culture, • Bring AI expertise to a massive real-world industry