AI Engineer — Agentic Systems & Retrieval Intelligence (BIM / Prefabrication)

MOD

MOD is an AI-native software platform to make industrialized construction 100x more intelligent, scalable, and efficient. For 95% of all construction projects globally the supply chain for prefabrication is not working. Property developers and architects struggle to check, validate, and integrate prefab elements early in their otherwise traditional construction projects.

MOD’s award-winning platform checks the prefab potential of any architectural design and orchestrates the supply chain to significantly increase the prefab share. It is 100% model-based and deeply integrated into the current design and procurement processes.

MOD is a category-defining AI company founded by serial entrepreneurs. MOD is closely linked to the world’s leading R&D centers for AI, digital fabrication, and architecture at ETH Zurich, TU Munich and Stanford.

A large part of MOD’s technology is developed as open source, including our domain language (mod-dlm) and validation engines. This is a core principle for us: by making our tools open and extensible, we accelerate adoption and attract a global community of contributors. For AI engineers, this means your work will shape both a high-impact product and the broader ecosystem: a unique opportunity to build systems that last.

WHAT IS YOUR ROLE

We are hiring a senior AI engineer to lead MOD’s agentic + retrieval intelligence layer. You will design workflows powered by ADK (Agent Development Kit), MCP (Model Context Protocol), and LangGraph, build RAG pipelines over BIM and supplier data, and develop geometry-aware reasoning models for prefabrication analysis and supplier-product matching. This is a founding team role with significant ownership and impact.

WHAT YOU DO

• Architect and implement agent workflows with ADK, MCP, and LangGraph for prefab analysis, validation, and supplier matching.
• Build domain-specific RAG systems using pgvector/FAISS with BIM embeddings and supplier metadata.
• Develop multi-modal retrieval combining text, geometry, logistics, and sustainability data.
• Prototype and train graph-based ML models (GNNs) and geometric embeddings for prefab suitability and supplier-product scoring.
• Integrate AI into mod-val (validation engine) and mod-rec (future model adaptation engine).
• Collaborate with BIM engineers on IFC.js, Speckle, Revit/Rhino integrations.
• Contribute to MOD’s open-source stack (mod-dlm, mod-val, mod-rec).

WHAT YOU BRING

Required:

• Strong in Python and LLM orchestration (ADK, MCP, LangGraph/LangChain).
• Hands-on with RAG + vector search (pgvector, FAISS, Weaviate).
• Experience with LLM tool-use, structured outputs, and function calling.
• Familiarity with BIM standards (IFC, Revit, Speckle) or ability to learn quickly.
• Track record of shipping AI systems into production (FastAPI, Docker, CI/CD, AWS/GCP).

Nice-to-have:

• Experience with geometric/graph ML (PyTorch Geometric, DGL).
• Knowledge of construction tech, DfMA, or prefabrication workflows.
• Contributions to open-source projects in AI or AEC domains.
• Publications or patents in retrieval, agentic AI, or 3D reasoning.

WHY JOIN

• Be the founding AI engineer with ownership over MOD’s intelligence layer.
• Build AI that directly reduces CO₂ emissions and transforms one of the largest industries on the globe.
• Opportunity to grow into Head of AI / Research Lead as the team scales.
• Collaborate with a deeply technical founding team and an active open-source ecosystem.
• Contribute to open-source projects that will become industry standards, and work in the open where impact extends far beyond MOD itself.

If you’re excited about building smart systems that think like engineers and architects — not just chatbots — apply now and help us reshape construction’s intelligence layer.