AI Engineer - Assistant Capabilities

About the company

Build is a New York–based AI infrastructure company that automates due diligence and development workflows for large infrastructure and data center projects.

About this role

ABOUT BUILD Build is creating the agentic AI stack for the built world. We help institutional real estate teams automate complex development and acquisitions workflows so important projects can move from concept to completion faster, with less cost, delay, and operational drag. Our customers include some of the largest built-world institutions: alternative asset investors, developers, infrastructure owners, energy companies, industrial operators, and public-sector partners. Their work shapes the physical world, but the workflows behind that work are still slow, fragmented, document-heavy, and dependent on expert coordination. We believe the next generation of built-world software will not just organize work. It will help do the work. Agents will reason across documents, drawings, financial models, market data, approvals, constraints, and expert judgment. Human experts will stay in control, but they will operate with far more leverage. We are backed by leading investors and operators, including executives from Blackstone and OpenAI, alongside top venture firms. We are building a generational company at the intersection of AI and the physical world. ABOUT THE ROLE We are looking for an AI engineer, product to build agentic workflows that customers use in production. This is a hands-on engineering role for someone who wants to turn ambiguous, high-value real estate workflows into reliable product experiences. You will work directly with customers, domain experts, designers, and engineers to understand how work actually happens, then build AI-powered systems that help experts move faster and make better decisions. You should be excited by the messy middle between product engineering and applied AI: long-running agents, context engineering, document intelligence, visual reasoning, workflow orchestration, human-in-the-loop review, evals, observability, and product surfaces that make agent work understandable and trustworthy. This is not a research role. It is also not a demo-building role. You will ship production software, own real customer outcomes, and help define what great AI-native product engineering looks like in the built world. WHAT YOU’LL DO - Design and ship production AI workflows for real estate design, acquisitions, diligence, planning, and project execution. - Work directly with customers and internal experts to map complex workflows into software systems with clear inputs, outputs, edge cases, and success criteria. - Build agents that reason across leases, zoning documents, site plans, drawings, financial models, market comps, investment memos, permits, emails, and project history. - Own full-stack product features end to end, from backend workflow logic to user-facing review, approval, and collaboration surfaces. - Design context strategies that help agents use the right documents, prior decisions, domain constraints, tool outputs, and user intent at the right time. - Build retrieval, extraction, structured-output, and tool-calling flows that are robust enough for expert users. - Create evals for document understanding, grounded reasoning, workflow completion, visual QA, accuracy, latency, and customer usefulness. - Inspect traces, debug failures, improve prompts and workflows, and turn customer feedback into measurable system improvements. - Partner with the core AI and infrastructure team to improve agent reliability, observability, evaluation, and developer velocity. - Raise the product quality bar for AI systems that need to be inspectable, explainable, and trusted. EXAMPLE PROJECTS - Build an acquisition diligence workflow that reads leases, zoning documents, market comps, financial models, and investment memos, then produces a grounded risk summary with citations and expert review. - Build a visual reasoning workflow that compares site plans, drawings, maps, and project constraints to identify feasibility issues. - Build a long-running development workflow agent that tracks missing documents, asks follow-up questions, routes work to the right expert, and prepares auditable recommendations. - Build a customer-facing review surface where experts can inspect agent reasoning, approve outputs, correct mistakes, and teach the system over time. - Build an internal expert workflow that turns repeated customer implementation work into reusable product primitives. WHAT SUCCESS LOOKS LIKE In your first few months, you will ship at least one production agent workflow used by customers or internal experts. It will have clear task success metrics, eval coverage, traceability, documented failure modes, cost and latency targets, and a feedback loop that makes the system better over time. Over the longer term, your work will help Build move from workflow automation to increasingly autonomous execution across the built world. YOU MAY BE A FIT IF - You are a strong product-minded software engineer who has shipped production systems used by real users. - You have built with LLM APIs, agent frameworks, structured outputs, tool calling, RAG, document processing, or workflow systems. - You are comfortable with Python and modern backend systems, and you can move across the stack when needed. - You can take a vague customer problem, ask the right questions, identify the workflow, and turn it into a product that works. - You care about agent quality beyond prompts: evals, traces, regressions, edge cases, latency, cost, and user trust. - You like working with domain experts and learning the details of complex industries. - You move fast, but you do not confuse a compelling demo with a reliable production system. - You have good product taste and care about making complex AI behavior understandable to users. BONUS POINTS - Experience with LangGraph, LangChain, Temporal, workflow engines, vector databases, reranking, document AI, multimodal models, or agent observability tools. - Experience building AI products in real es