2026 PhD Residency, Machine Learning Explainability (Tapestry)

About the role

As a PhD Resident in Machine Learning Explainability, you will join Tapestry’s six-month PhD Residency Program to research how modern AI techniques—particularly large language models (LLMs) and graph-based models—can explain complex decision-making systems used in electric grid operations.

This role focuses on improving trust, transparency, and usability of highly constrained optimization engines such as economic dispatch, unit commitment, and long-term planning tools. Your work will explore how complex model outputs, constraints, and system behaviors can be translated into clear, human-understandable explanations for both expert and non-expert users.

How you will make 10X Impact

• Research and prototype explainability approaches for complex optimization and decision-making systems in the electric grid.
• Apply LLMs to generate natural-language explanations of model outputs, constraints, and tradeoffs.
• Explore reasoning over graph-structured data (e.g., power grids) to produce grounded, faithful explanations.
• Investigate methods to detect and explain incorrect or anomalous network model inputs in natural language.
• Translate academic research into practical feasibility demonstrations using real or simulated grid data.
• Collaborate with machine learning researchers and power systems experts to refine approaches and evaluation methods.
• Clearly document findings and communicate insights to inform future research and product directions.

What you should have:

• Currently enrolled in a PhD program in Machine Learning, Computer Science, Electrical Engineering, or a related field.
• Strong research experience in machine learning or deep learning.
• Hands-on experience with LLMs, transformer-based models, or graph neural networks.
• Strong programming skills in Python and experience with frameworks such as PyTorch or JAX.
• Ability to reason about complex systems and communicate technical concepts clearly.
• Interest in applying AI to high-impact, real-world infrastructure challenges.

It’d be great if you also had one or more of these:

• Experience with optimization, convex optimization, or decision-making systems.
• Familiarity with power systems, energy modeling, or networked physical systems.
• Prior work in explainability, interpretability, or human-centered AI.
• Publications in ML or AI venues (e.g., NeurIPS, ICML, ICLR).
• Experience working on applied research projects in industry or startup environments.