ssm-foundations
in progressA foundations book bridging numerical analysis and dynamical systems to modern sequence models, live and in progress.
ssm-foundations.brandon-behring.dev →Statistical modeling for insurance, built with applied-math rigor.
Applied-math PhD with four years of production modeling at Prudential, now looking for statistical modeling work in insurance. I build causal-inference and AI-evaluation tooling, foundations books on sequence models and reinforcement learning, and the research infrastructure beneath them.
Rigor is the fulcrum; AI is the lever.
Applied-math PhD (NJIT 2020; NYU postdoc) · peer-reviewed publications · four years in industry
Selected work
double_ml_time_series
in progressDouble machine learning extended to time series: cross-fitting that respects time ordering, with a live web edition.
dml.brandon-behring.dev →A research study: do prompt-injection detectors generalize to attack families they were never trained on?
Deployed study →DeepLearning.AI Study Notes
in progressA searchable, math-rendered corpus of study notes synthesized from DeepLearning.AI short courses, with original practice questions and spaced-recall flashcards.
study-notes.brandon-behring.dev →Interactive citation graph of the RL + control literature: drag, search, and filter by theme.
from: rl_and_control: RL + Control Theory guide →An interactive eigenvalue explorer: watch a system's stability region shift as the discretization step changes.
from: ssm-foundations →Built to be audited
AI expands what one person can ship; rigor decides what is allowed to ship. Substantive artifacts carry a decision log and an independent audit, and key claims trace back to source: the math is hand-derived and checkable.
About
I work on modern sequence-model architectures (state-space models such as Mamba and S4) and on reinforcement learning with optimal control, and I turn that work into rigorous, auditable books and tooling. Applied causal inference, risk analysis, and evaluation are the methodology spine throughout.
Practitioner guides for data, ML & AI engineering. Read the guides →
Before industry: peer-reviewed research in applied mathematics, on point-vortex and Hamiltonian dynamics and on network epidemiology. Publications →
Contact
- Email: brandon.m.behring@gmail.com
- GitHub: @brandon-behring
- LinkedIn: /in/brandon-behring
Rigor is the fulcrum; AI is the lever.