Data Scientist – Building Energy Modeling & Machine Learning
Employment Type:
Full-Time
About Us:
We are a fast-growing company developing next-generation economic and energy analysis software. Our mission is to make building performance evaluation faster, smarter, and more reliable through the combined power of building science and artificial intelligence.
We’re seeking a Data Scientist to lead the development of machine learning models that predict simulated building energy use and support key features in our analysis platform. You’ll work closely with our engineering and building science teams to design, validate, and deploy data-driven solutions that bridge simulation and real-world performance.
We’re seeking a Data Scientist to lead the development of machine learning models that predict simulated building energy use and support key features in our analysis platform. You’ll work closely with our engineering and building science teams to design, validate, and deploy data-driven solutions that bridge simulation and real-world performance.
What You’ll Do:
● Translate use cases into well-posed ML/statistical tasks with clear evaluation metrics and traceable decisions.
● Collaborate with building scientists to design and implement evaluations grounded in physical building models.
● Design, train, and deploy models for fault attribution, asset/entity normalization, code/spec parsing, document-grounded QA (RAG), and structured outputs with tool-use.
● Build robustness tests (data-slice/shift, noise injection, OOD detection), quantify uncertainty and calibration, and develop red-teaming and failure mode taxonomies.
● Leverage EnergyPlus, Comstock, and other simulation testbeds for control-policy benchmarking and synthetic data generation, with clear gap analysis to real-world operations.
● Manage the model lifecycle: experiment tracking, model registry (e.g., MLflow), feature serving, lineage, versioning, and audit logs.
● Publish technical write-ups or papers, and contribute to open-source initiatives.
● Collaborate with building scientists to design and implement evaluations grounded in physical building models.
● Design, train, and deploy models for fault attribution, asset/entity normalization, code/spec parsing, document-grounded QA (RAG), and structured outputs with tool-use.
● Build robustness tests (data-slice/shift, noise injection, OOD detection), quantify uncertainty and calibration, and develop red-teaming and failure mode taxonomies.
● Leverage EnergyPlus, Comstock, and other simulation testbeds for control-policy benchmarking and synthetic data generation, with clear gap analysis to real-world operations.
● Manage the model lifecycle: experiment tracking, model registry (e.g., MLflow), feature serving, lineage, versioning, and audit logs.
● Publish technical write-ups or papers, and contribute to open-source initiatives.
Minimum Qualifications:
● Education & Experience: Master’s degree with at least 3 years of full-time experience in Building Science and Machine Learning, or a closely related field.
● Machine Learning Expertise: Proven experience designing, developing, and deploying advanced ML/DL models, including structured prediction, representation learning, and out-of-distribution (OOD) generalization.
● Building Science Knowledge: Strong understanding of HVAC systems, control logic, energy performance, and building operations.
● Model Reliability & Robustness: Proven ability to implement robustness testing, uncertainty estimation, and calibrated performance reporting.
● Statistical Foundation: Strong foundation in designing and analyzing experiments, including hypothesis testing, resampling methods, error analysis, and ensuring data integrity by controlling for bias and leakage.
● LLM Workflows: Experience working with LLM-based workflows, including evaluating retrieval quality, grounding model outputs, and checking for hallucinations or factual accuracy.
● Model Development & Tracking: Experience with Git-based version control and experiment tracking tools such as MLflow and Neptune.ai for managing and monitoring reproducible ML pipelines.
● Machine Learning Expertise: Proven experience designing, developing, and deploying advanced ML/DL models, including structured prediction, representation learning, and out-of-distribution (OOD) generalization.
● Building Science Knowledge: Strong understanding of HVAC systems, control logic, energy performance, and building operations.
● Model Reliability & Robustness: Proven ability to implement robustness testing, uncertainty estimation, and calibrated performance reporting.
● Statistical Foundation: Strong foundation in designing and analyzing experiments, including hypothesis testing, resampling methods, error analysis, and ensuring data integrity by controlling for bias and leakage.
● LLM Workflows: Experience working with LLM-based workflows, including evaluating retrieval quality, grounding model outputs, and checking for hallucinations or factual accuracy.
● Model Development & Tracking: Experience with Git-based version control and experiment tracking tools such as MLflow and Neptune.ai for managing and monitoring reproducible ML pipelines.
Preferred Skill Sets
● Experience with EnergyPlus and OpenStudio for building energy simulation and performance benchmarking.
● Experience with explainability techniques to quantify feature contributions and improve model transparency.
● Familiarity with cutting-edge neural network architectures and their application to structured or scientific datasets.
● Experience running large-scale simulations or training models on cloud-based environments.
● Experience with explainability techniques to quantify feature contributions and improve model transparency.
● Familiarity with cutting-edge neural network architectures and their application to structured or scientific datasets.
● Experience running large-scale simulations or training models on cloud-based environments.
Preferred Skill Sets
● Work at the cutting edge of building energy science and machine learning.
● Collaborate with a passionate, cross-disciplinary team of engineers and building scientists.
● Competitive compensation with equity opportunities.
● Flexible, fully remote work environment.
● Collaborate with a passionate, cross-disciplinary team of engineers and building scientists.
● Competitive compensation with equity opportunities.
● Flexible, fully remote work environment.
