Backend Engineer - Agent Builder: Revit Integration
What we do:
Architects and engineers collaborate with our agents to design & engineer buildings 10x faster.
- First, we developed a system that transforms BIM data (the source code of buildings) into an AI-reasonable format.
- Now, we offer agents in SaaS + Service that build buildings with architects and engineers.
Tech Stack:
- .NET 8.0 / ASP.NET Core
- SignalR (real-time communication)
- RabbitMQ (message queuing)
- Entity Framework Core
- Revit API / Autodesk Platform Services
- MCP (Model Context Protocol)
- Docker/AWS ECS
What Need to be built
-
Model Context Extractor (Revit Add-in)Serialize current model state to JSON
Send relevant context with each prompt
-
C# Code Generator Prompt EngineeringTemplate system for Claude
Include Revit API constraints
Transaction handling patterns
-
Dynamic Code Executor (Revit Add-in)Roslyn compiler integration
Safe execution environment
Result serialization
-
Session ManagerTrack Desktop App ↔ Revit pairings
Handle disconnections gracefully
Queue commands for offline Revit
Key Requirements:
- Proficiency in C# and .NET 8.0, with experience building scalable backend services using ASP.NET Core
- Strong understanding of message-driven architecture, including hands-on experience with RabbitMQ or similar queuing systems
- Experience with Entity Framework Core and working with relational databases in production environments
- Familiarity with real-time communication frameworks, particularly SignalR, for client-server messaging
- Demonstrated ability to build or integrate with Autodesk Revit API or similar CAD/BIM platforms
- Comfortable working with dynamic code execution and sandboxing, ideally using tools like Roslyn
- Experience developing or consuming add-ins/plugins for desktop software (e.g., Revit, AutoCAD, etc.)
- Exposure to cloud infrastructure, especially Docker and AWS ECS, for containerized deployments
- Strong software architecture skills—able to build reliable, event-driven systems with decoupled services
- Ability to work in fast-paced, experimental environments involving AI integrations and prompt engineering
Bonus: Experience with Claude, LLM-based code generation, or AI-powered developer tools