About the Role
Our Client’s mission is to empower people to design and build the infrastructure that supports life on Earth and beyond. They are focused on enhancing human potential by giving both technical and non-technical users the ability to design, estimate costs, and plan the schedules of complex construction projects.
In this role, you will contribute directly to our work on Graph Neural Networks (GNNs). You will help design architectures and modules, work extensively with graph structures, and develop methods to identify, represent, and modify elements, properties, and hierarchies within complex systems. You will be part of the core GNN team, collaborating closely with other groups to push the boundaries of human–machine collaboration.
If you are driven by ambitious missions, love tackling tough technical challenges, and are not afraid to explore bold ideas (and fail often in the process), this opportunity is for you.
Key Responsibilities
- Develop and implement advanced Graph Neural Network (GNN) models to analyze, verify, and modify complex structural assemblies and component representations within detailed 3D models—covering domains such as architecture, aerospace, automotive, and industrial design.
- Design and optimize graph-based data structures and computer vision (CV) models for identifying, representing, and modifying 3D meshes, topologies, and hierarchical relationships.
- Evaluate and enhance the impact of computational graph representations on execution efficiency and model performance across diverse hardware platforms (e.g., GPUs, CPUs).
- Explore and apply methods for continual and self-supervised learning using 3D data, enabling adaptive AI systems that evolve and improve over time.
- Conduct research and experimentation with emerging technologies in graph learning and AI, ensuring the team’s solutions remain cutting-edge and future ready.
Requirements
- 5+ years of relevant experience or a portfolio showcasing technically impressive projects.
- Strong expertise in Graph Neural Network (GNN) frameworks, architectures, and applications.
- Hands-on experience with Reinforcement Learning (RL).
- Solid understanding of 3D environments, geometry, and spatial data.
- Proficiency in data handling, database management, and large-scale data processing.
- Familiarity with knowledge graph technologies such as Neo4j, TigerGraph, or Amazon Neptune, and related semantic web standards (OWL, RDF, SWRL, SPARQL, JSON-LD) is a significant plus.
- Comfort working with computational graphs and intermediate representations (compiler or systems background is a strong advantage).
- Deep knowledge of modern vector and embedding-based graph representation techniques, including experience designing and implementing algorithms for their generation.
- Collaborative mindset with the ability to work effectively across interdisciplinary teams.
- Quick learner with a passion for continuous growth and innovation.
Good to Have
- Experience working with Large Language Models (LLMs), convex optimization techniques, and physics-based simulation.
- Strong background in developing and deploying graph analytics solutions within distributed systems, including tasks such as community detection, pattern recognition, subgraph extraction, and importance or influence ranking.
- Domain experience in architecture, aerospace, or automotive industries is highly desirable.
