Phd Researcher, Machine Learning for Construction

Autodesk•Boston, MA

1d•$118,560 - $162,240•Hybrid

About The Position

Autodesk Research is looking for a PhD Researcher Intern in machine learning to help advance AI for the architecture, engineering, and construction (AEC) industry. AEC projects generate large amounts of data across design, planning, and construction. Our team studies how machine learning and AI can connect this data and learn from them to understand, predict, and reason about how projects unfold. As an Researcher intern, you will own a research problem in this area, scoped to fit your expertise and the team's current direction. You'll collaborate with researchers and industry domain experts, and your work will help shape how Autodesk brings AI to AEC workflows.

Requirements

Currently pursuing a PhD in Computer Science, Machine Learning, Data Science, Civil Engineering, Construction Informatics, or a related field with a graduation date no sooner than April 2027.
Research experience in graph machine learning on heterogeneous knowledge graphs, including graph neural networks, graph transformers, link prediction, node classification, and graph self-supervised learning (e.g., masked graph autoencoders, joint-embedding predictive architectures (JEPA)).
Experience with temporal or time-varying data, through temporal graph models or time-series forecasting.
Experience with uncertainty-aware prediction, including calibrated probability estimates and learning from small or imbalanced labeled data.
Rigorous experimental practice: baselines, ablations, leakage-safe train/test splits, and label quality checks.
Proficiency in Python and PyTorch (PyTorch Geometric or DGL a plus), and graph query languages.
Self-directed, with the ability to explain technical results clearly to both technical and non-technical audiences.

Nice To Haves

Time-series, state-space, or multimodal modeling, including combining time-varying signals from multiple sources and modalities.
Bayesian or probabilistic modeling.
Natural language processing or large language models for classifying or extracting information from technical text and AEC data.
Experience with AEC data or workflows, such as building models, drawings, project schedules, project records, or design review, etc.
Knowledge graph engineering: Neo4j/Cypher, ontologies, entity linking, and data integration.
Publications at top machine learning venues (e.g., NeurIPS, ICML, KDD) or construction computing venues (e.g., Automation in Construction, ASCE computing journals, ISARC, CIB W78).

Responsibilities

Define and investigate an open research question in machine learning and AI for AEC data, from problem framing through experiments and results.
Design and train AI and machine learning models, including graph-based and temporal models, on large, heterogeneous real-world project data.
Build predictive models with reliable uncertainty estimates, and evaluate them rigorously against strong baselines.
Prepare, integrate, and validate data from multiple sources, including knowledge graph representations.
Document methods and results, present findings, and potentially contribute to research publications.