Senior Staff AI Engineer

📋 Description

• • Own the end-to-end technical vision for Homebound’s vertical AI stack, translating the complexity of residential construction into elegant, LLM-powered solutions that delight customers and drive margin.
• • Architect and ship production-grade systems that ingest millions of data points—from geotechnical reports to supply-chain invoices—and surface actionable insights to construction managers, designers, and homeowners in real time.
• • Lead the design and deployment of novel RAG pipelines, agentic workflows, and eval frameworks that continuously improve the accuracy, safety, and cost-efficiency of our GenAI platform.
• • Establish rigorous experimentation standards (A/B, offline evals, human-in-the-loop) so every new model or prompt change is measurable, reversible, and aligned with business KPIs such as cycle-time reduction, cost per square foot, and Net Promoter Score.
• • Write clean, scalable Python services deployed on Kubernetes, instrumented with OpenTelemetry, and monitored with custom dashboards that surface drift, latency, and token spend to any stakeholder in under 30 seconds.
• • Collaborate daily with product managers, field superintendents, and UX designers to translate on-the-ground construction pain points into AI features—whether that’s an LLM that auto-generates RFIs or a computer-vision model that flags drywall defects from phone-camera images.
• • Mentor a growing team of ML and backend engineers through architecture reviews, pair programming, and internal tech talks; raise the bar for code quality, testing, and MLOps across the entire engineering org.
• • Champion responsible AI practices—bias audits, privacy safeguards, and transparent model cards—ensuring our technology earns homeowner trust and meets regulatory requirements in every state we operate.
• • Stay on the bleeding edge: prototype new foundation models, evaluate open-source alternatives, and present findings to exec staff so Homebound can adopt breakthrough capabilities ahead of the market.
• • Drive cross-functional OKRs that link model improvements directly to financial outcomes, such as shaving 10 days off build schedules or cutting material waste by 5 %, and present quarterly impact reports to the board.
• • Contribute to Homebound’s patent portfolio by identifying novel applications of AI in construction and partnering with legal to secure IP that widens our competitive moat.
• • Foster a culture of psychological safety and intellectual curiosity where every engineer—from intern to principal—feels empowered to question assumptions and propose bold ideas.