Founding AI Engineer

About the Role
  
A fast-growing technology company is developing an AI-native platform focused on transforming how buildings and physical spaces are designed, estimated, and delivered. The company is moving away from manual construction and modelling workflows by using AI to automate the generation of building models from raw 2D drawings and spatial data.
As a Founding AI Engineer within the core AI and 3D team, you would take ownership of the full pipeline, from raw spatial or scan data through to structured, semantically rich digital building models. This is not a traditional BIM modelling position; it is a highly technical systems and algorithms role focused on creating the infrastructure that automates and replaces manual BIM workflows.
  
Key Responsibilities

• Own the end-to-end development of the scan-to-model pipeline, including data ingestion, registration, semantic segmentation, geometric analysis, and automated model generation. You will be responsible for ensuring the reliability and accuracy of the system from input through to final output.
• Develop and improve large-scale 3D scene understanding algorithms capable of identifying, segmenting, and reconstructing building elements such as walls, floors, ceilings, and MEP systems from noisy and incomplete real-world scan data. The focus is on solving practical production challenges rather than purely academic benchmark problems.
• Convert AI-generated outputs into industry-standard, interoperable building models compatible with widely used AEC workflows and software platforms. This includes bridging the gap between machine learning predictions and production-ready digital building assets that can be used reliably by design and construction teams.
• Define and implement evaluation frameworks, validation pipelines, and quality metrics to measure not only technical model accuracy, but also constructability, usability, and real-world performance. Where suitable standards or benchmarks do not yet exist, you will help establish them.

Requirements

• Proven experience working with 3D computer vision, LiDAR, point cloud processing, or large-scale geometric reconstruction systems in real-world production environments.
• Strong programming skills in Python and/or C++, with the ability to design, build, and deploy complete end-to-end pipelines rather than contributing only to isolated components.
• Working knowledge of 3D processing frameworks such as Open3D, PCL, or similar technologies used for spatial data analysis and reconstruction.
• Comfortable operating in fast-moving and ambiguous technical environments, with a strong first-principles approach to problem-solving and system design.
• Experience developing semantic or instance segmentation models for 3D spatial data, including the classification of structural building elements such as walls, floors, and ceilings.
• Exposure to graph neural networks or machine learning techniques designed for 3D or graph-structured datasets.
• Background working with digital twin technologies, reality capture platforms, or building reconstruction systems.
• Experience with procedural or parametric model generation workflows.
• Familiarity integrating AI-generated outputs into CAD, BIM, or related design and construction software environments.