Senior Software Engineer with AI plus Revit

General Description

The Senior Software Engineer with AI plus Revit builds the customer-facing AI layer for Stratus — the production agent systems, tool integrations, and evaluation infrastructure that let our products reason over MEP fabrication data and act on a contractor's behalf. This is a hands-on senior engineering role: you will design and ship multi-agent workflows, build the tool and context layer that connects agents to Stratus data, and own the guardrails, evals, and observability that make those systems safe and trustworthy in front of customers.

Key Responsibilities

• AI engineering (primary): Design and build customer-facing agentic workflows: multi-agent orchestration (e.g., LangGraph, CrewAI, AutoGen), tool calling, structured outputs, multi-step planning, and human-in-the-loop checkpoints, to automate complex MEP engineering tasks.
• Evaluation, guardrails, and failure-mode analysis for agent systems — including offline eval suites in CI and live production sampling — to ensure they are safe, reliable, and grounded.
• Build the tool and context layer connecting agents to Stratus data via internal and customer-facing APIs (e.g., MCP), including context-window management, permissioning, and cost control.
• Set up observability and tracing for agent behavior; diagnose cost, latency, and hallucination issues in production.
• Integrate Stratus's published design and fabrication data into agent workflows.

Revit Integration

• Build and maintain the Revit-side integrations that feed the AI layer — the add-ins and publishing paths that move design and fabrication data out of our customers' Autodesk Revit environment and into Stratus.
• Write production C#/.NET code against the Revit API: custom commands, external events, document and transaction management, and integration with Autodesk model data.
• Own the fidelity, mapping, validation, and error handling that keep Revit data exports trustworthy enough for agents to reason over.
• Solve the hard problems specific to Revit add-in development — version compatibility, performance inside large models, the API threading model, and graceful degradation when the host environment misbehaves.

Collaboration & Delivery

• Use AI-assisted development tooling (Claude Code, Cursor, Copilot, etc.) as a first-class part of the dev loop — writing tests for AI-generated changes and exercising clear judgment about when AI output is ready to ship.
• Collaborate with product managers, designers, and customer-facing teams to scope, design, and ship — grounding technical decisions in real design and fabrication workflows.
• Contribute to agile workflows, ensuring flexibility and responsiveness to evolving project needs.
• Share knowledge and raise the engineering bar through code review and pragmatic best practices.

Qualifications

• 6+ years of software engineering experience, with a proven track record of shipping and operating production-grade systems — not just prototypes or notebooks.
• Primary — AI engineering: Hands-on experience building and operating customer-facing agentic systems in production — orchestration frameworks (LangGraph, CrewAI, or AutoGen), tool calling, structured outputs, and eval frameworks. Experience with evals, guardrails, and observability for LLM or agent systems.
• Secondary — Revit API: Hands-on production experience building Revit add-ins and working within the realities of the Revit API — the document and transaction lifecycle, external events, the threading model, version compatibility, and performance inside large models. Strong proficiency in C#/.NET, with demonstrated production ownership of real features. Experience with MCP or similar tool-integration protocols. Strong computer science fundamentals (data structures, algorithms, system design) and solid API/backend engineering depth.
• Hands-on use of AI-assisted development tooling (Claude Code, Cursor, Copilot, or equivalent) as a first-class part of your daily workflow, with clear judgment about when AI output ships, needs rework, or should be thrown away.
• Excellent communication skills — able to explain complex AI systems clearly to teammates, product partners, and customers. Comfort working in newly forming, ambiguous areas where learning and adaptability are key.
• Degree in Computer Science, Engineering, or a related field, or equivalent practical experience.

Nice to Have

• Familiarity working with data derived from CAD/BIM or other 2D/3D model sets.
• Working familiarity with the broader Autodesk application family (e.g., AutoCAD, Fabrication, BIM 360 / ACC, Navisworks) and the Autodesk Platform Services ecosystem (formerly Forge) — Data Management, Model Derivative, Design Automation, or related APIs.
• Desktop application development experience — including installers (MSIs) and managing packaging, deployment, and updates across customer environments.
• Familiarity with machine learning concepts and how data is represented for training.
• Domain knowledge of MEP, BIM, or construction fabrication workflows, or an architectural/AEC engineering background.
• Full-stack development experience beyond the desktop application.
• Background in customer-facing or professional-services roles.
• Drive to continually learn new technologies and seek new ways to solve hard problems.
• Bias toward putting your ideas out there and failing fast.