cv
Contact Information
| Name | Gabriel Earle |
| gabrieljearle@gmail.com | |
| Location | Austin, TX |
Professional Summary
PhD mathematician and AI engineer specializing in the design of production-oriented AI systems. Expert architect of end-to-end AI platforms combining large language models, agentic workflows, and knowledge engines to perform automated reasoning and domain-specific decision support. Research areas include transformer architectures, machine learning, statistical modeling, numerical methods, and scientific computing.
Education
-
2014 – 2020 PhD in Mathematics
Duke University
- Thesis: "Stratified mcmc sampling of non-reversible dynamics"
- Supervisor: Jonathan Mattingly
-
2010 – 2014 Bachelor of Science in Mathematics
University of Texas, Austin
Experience
-
2025 – Present Austin, Tx
Founder and AI Engineer
Constructed Mind LLC
- Built an end-to-end AI system combining large language models, agentic workflows, knowledge graphs, retrieval-augmented generation, and multimodal reasoning to automate building-code compliance analysis.
- Designed a knowledge-engineering platform capable of translating regulatory language into executable reasoning and verification procedures. - Developed graph-based semantic representations, ontology generation pipelines, retrieval systems, and autonomous AI agents for technical decision support.
-
2021 – 2025 Amherst, MA
Visiting Assistant Professor, Department of Mathematics & Statistics
University of Massachusetts Amherst
- Investigated transformer architectures and open-source large language models, developing a strong understanding of modern machine learning systems, neural networks, and AI model implementation.
- Conducted research in statistical modeling, stochastic systems, and scientific computing, developing novel sampling algorithms for high-dimensional simulation and rare-event analysis.
- Mentored undergraduate researchers and taught university mathematics courses, communicating complex technical concepts to diverse audiences.
-
2014 – 2021 Durham, NC
Ph.D. Candidate & Post-Doctoral Researcher, Department of Mathematics
Duke University
- Formed research projects and developed interests by connecting with mentors and joining groups with peers in mathematics
- Presented posters and talks on research at major conferences across the nation, building connections with top authorities in field of study
- Led organization efforts for departmental research group, focused on sharing individual and group research results in an informal setting for the purpose of developing collaborative relationships across departments
Selected Publications
-
2025 Relative Entropy Methods for the Approximation of Reactive Trajectories and Committor Functions
Gabriel Earle, Brian Van Koten — SIAM/ASA Journal on Uncertainty Quantification
-
2023 Aggregation methods for computing steady states in statistical physics
Gabriel Earle, Brian Van Koten — Multiscale Modeling & Simulation
-
Aug 2021 Convergence of stratified MCMC sampling of non-reversible dynamics
Gabriel Earle, Jonathan Mattingly — Partial Differential Equations: Analysis and Computations
Skills
AI Systems and Machine Learning: LLMs, Agentic AI, RAG, Fine-Tuning, Transformers, Deep Learning, Generative AI, Neural Networks, Model Quantization, Technical Strategy
Knowledge Engineering: Knowledge Graphs, Ontologies, Semantic Search, Neo4j, Vector Databases
Mathematics & Modeling: Probability, Statistics, Statistical Modeling, Numerical Analysis, Optimization, Scientific Computing
Programming: Python, PyTorch, TensorFlow, Julia, Java, C++, MATLAB
Interests
Creative Writing: Science Fiction, Fantasy, Speculative Fiction
History and Culture: Ancient History, Military History
Music: All Genres