Computational Scientist, Senior

Booz Allen Hamilton•Washington, DC

About The Position

As a computational scientist, you operate at the intersection of mathematics and software, developing systems where code correctness relies on the underlying scientific principles. This role involves creating numerical methods, implementing surrogate models, or building simulation frameworks that require a deep understanding of both physics and architecture. Booz Allen is developing a platform designed to make computational models composable, verifiable, and fast enough for real-time operational decisions. As the scientific computing lead, you will be responsible for the foundational layer supporting two integrated products. One product demands models that can be independently validated, checked for compatibility when combined, and executed with traceable evidence of correctness. Your role will involve designing the validation logic, compatibility checking, and observable binding to ensure trustworthy composition. The second product requires probabilistic estimation from noisy, incomplete data streams, where you will build the uncertainty modeling and state estimation to transform raw signals into a functional understanding of the world for a solver. Your work will form the critical foundation for both products. You will join a small, expert team, operating at the technical forefront, and be responsible for transforming technical concepts from research papers or whiteboards into tested, production-grade code. The work will be conducted in a rapid development environment utilizing AI-native tooling, ensuring scientific rigor at engineering speed.

Requirements

5+ years of experience developing computational models, numerical solvers, or physics-based simulation codes in a professional or research environment
Experience implementing reduced-order modeling techniques, surrogate modeling, or data-driven approximation methods for dynamical systems
Experience defining and enforcing validity conditions for computational models, including parameter bounds, domain constraints, or interface specifications between heterogeneous codes
Knowledge of probability theory, statistical estimation, or uncertainty quantification as applied in computational systems
Ability to translate mathematical formulations from research literature into tested, maintainable production software
Ability to obtain a TS/SCI clearance with a polygraph
Bachelor's degree in an Applied Mathematics, Physics, Computational Science, Mechanical Engineering, or Aerospace Engineering field, or 8+ years of experience working in a professional environment in lieu of a degree

Nice To Haves

Experience with operator-theoretic methods such as Koopman analysis, dynamic mode decomposition, or spectral methods for nonlinear systems
Experience building model repositories, simulation frameworks, or workflow systems that manage heterogeneous computational assets
Experience building products for defense, intelligence, or national security customers
Master's degree in Applied Mathematics, Physics, Computational Science, or a related quantitative field preferred
Doctorate degree in Applied Mathematics, Physics, Computational Science, or a related quantitative field a plus

Responsibilities

Own the layer that underpins two integrated products.
Design the validation logic, compatibility checking, and observable binding that makes composition trustworthy for models that can be validated in isolation, checked for compatibility when composed together, and executed with traceable evidence of correctness.
Build the uncertainty modeling and state estimation that turns raw signals into a working picture of the world that a solver can reason over for probabilistic estimation from noisy, incomplete data streams.
Take a concept from a technical paper or a whiteboard and turn it into tested, production-grade code.
Build in a rapid development environment with AI-native tooling, delivering scientific rigor at engineering velocity.