AI Engineer – Reinforcement Agent / BIM Geometry Automation
Location
India / Remote / Hybrid
About the Role
We are building an AI-powered reinforcement agent that can generate, validate, and edit reinforcement for structural elements such as walls, slabs, beams, columns, and precast components.
This is not a generic chatbot or RAG role. The work sits at the intersection of:
• AI agents and structured reasoning
• Computational geometry and 3D/2D shape handling
• BIM / IFC workflows
• Rule-based engineering automation
• Reinforcement detailing logic
We are looking for an engineer who can help design and build a production-grade system that understands element geometry, engineering intent, detailing rules, and output constraints, and converts them into reinforcement layouts, shape definitions, and downstream deliverables such as BBS, IFC-ready outputs, and visual review workflows.
What You Will Work On
• Build an AI-assisted reinforcement generation engine for structural concrete elements.
• Convert geometry inputs into reinforcement intent, bar layouts, and deterministic outputs.
• Design systems that combine LLMs, rule engines, and geometric algorithms rather than relying only on prompting.
• Work with IFC/BIM data, element geometry, local coordinate systems, openings, drilling zones, covers, hooks, laps, bend radii, and edge conditions.
• Develop logic for reinforcement placement across walls, slabs, beams, columns, and precast elements.
• Build validation workflows to check engineering constraints, collisions, spacing, cover, continuity, and constructability.
• Generate structured outputs for downstream rendering, editing, and BBS workflows.
• Support visual review flows where users can inspect, edit, and approve generated reinforcement.
• Improve the agent’s reliability on edge cases through evals, test datasets, deterministic checks, and debugging tools.
• Collaborate closely with product, structural engineering, and frontend/viewer teams.
Key Responsibilities
• Design backend services and workflows for reinforcement-agent execution.
• Build geometry-aware processing pipelines from IFC / extracted element data.
• Implement reinforcement placement logic using structured engineering rules.
• Create schemas for engineering intent, bar sets, bar positioning, and output validation.
• Integrate LLM/agent components where they genuinely add value, such as reasoning over engineering context, exception handling, or workflow orchestration.
• Ensure outputs are deterministic, traceable, and debuggable.
• Build or integrate tools for bar visualization, review, and correction.
• Write tests for critical edge cases including openings, corners, peripheral bars, lifters, face mesh interaction, and shape generation.
• Contribute to production deployment, monitoring, and performance optimization.
Must-Have Skills
• Strong Python experience.
• Strong backend engineering experience with APIs and production systems.
• Experience building AI/agent systems using structured outputs, tool use, workflow orchestration, or LangGraph-like systems.
• Good understanding of computational geometry, coordinate systems, transformations, and rule-based logic.
• Experience handling structured engineering or CAD/BIM-like data.
• Ability to break complex domain logic into deterministic, testable modules.
• Comfort working with LLMs as one component in a larger system rather than as the entire solution.
• Strong debugging and systems thinking.
Strongly Preferred
• Experience with IFC, BIM, IfcOpenShell, CAD/CAM, Revit, Allplan, Tekla, or similar ecosystems.
• Experience with reinforcement detailing, rebar modeling, BBS generation, or structural engineering workflows.
• Understanding of concepts such as cover, spacing, hooks, bend radius, lap lengths, anchorage, edge bars, opening reinforcement, mesh placement, and constructability.
• Experience with 3D viewers or geometry visualization pipelines.
• Experience with OpenCV / OCR / drawing extraction for engineering drawings.
• Experience building review tooling, QC pipelines, or human-in-the-loop engineering systems.
Nice to Have
• Structural or civil engineering exposure.
• Experience with Three.js, WebGL, or browser-based model viewers.
• Experience with optimization and search in geometric layouts.
• Familiarity with shape-code systems, reinforcement catalogs, and regional detailing standards.
• Experience creating evaluation datasets and benchmark workflows for engineering AI systems.
What Success Looks Like
In this role, success means you can help us move from “interesting prototype” to a robust system that:
• Understands element geometry correctly.
• Places reinforcement using explicit engineering logic.
• Handles real-world edge cases.
• Produces structured, editable outputs.
• Supports visual review and correction.
• Improves reliably over time through tests, evals, and production feedback.
Who Will Be a Good Fit
You are likely a strong fit if you are one of these:
• An AI engineer with strong geometry/CAD/BIM instincts.
• A computational geometry or graphics engineer interested in engineering automation.
• A backend AI engineer who has worked on structured reasoning systems and can learn reinforcement logic fast.
• A technically strong engineer with exposure to structural detailing or BIM workflows.
Who Will Not Be a Good Fit
This role is likely not a fit for someone whose experience is limited to:
• Generic RAG/chatbot projects only.
• Prompt engineering without backend/system design.
• LLM demos without production ownership.
• AI search/retrieval roles with no geometry, CAD, or engineering logic exposure.
Tech Stack Context
Our work may involve parts of the following stack depending on the module:
• Python
• FastAPI
• LLM APIs / agent frameworks
• IFC / IfcOpenShell / geometry processing
• Postgres / structured schemas
• OpenCV / extraction pipelines
• Three.js or viewer integrations
• AWS / production cloud workflows
