Optimization Scientist

About QpiAI

QpiAI is a deep-tech company in AI, Optimization, HPC, and Quantum Computing, with the QpiAI-Opt platform delivering decision intelligence for critical industries.

Role Overview

Develop next-generation, high-performance optimization solvers, modeling systems, and advanced algorithmic frameworks for the QpiAI-Opt platform, solving large-scale, complex combinatorial and numerical optimization problems. Requires deep expertise in optimization theory, algorithm design, and scientific software development.

Key Responsibilities (Concise Summary)

• Solver Development: Design and implement state-of-the-art LP, MILP (simplex, interior-point, branch-and-cut), NLP (SQP, interior-point), combinatorial, and QUBO/Ising solvers.
• Modeling: Develop mathematical model abstraction layers, DSLs, and flexible solver interfaces.
• Algorithm Research: Implement and benchmark advanced Branch-and-Bound/Cut, cutting plane, decomposition (Lagrangian, Benders, column generation), and metaheuristic techniques.
• HPC & Architecture: Design parallel/distributed solver architectures, implement GPU acceleration, and utilize HPC strategies (MPI).
• Performance: Develop rigorous benchmarking, profiling tools, and contribute to research publications and IP.

Required Qualifications

• Education: Masters or PhD in Mathematics & Computing, Mathematical Optimization, Operations Research, Applied Mathematics, or a related quantitative field.
• Theoretical Foundation: Strong foundation in Linear, Integer, Convex, and Numerical Optimization, plus graph/combinatorial algorithms.
• Experience: Hands-on experience with commercial or open-source solvers.
• Programming: Strong proficiency in C++, Python, Julia, or Rust for performance.

Preferred Qualifications

Prior experience with optimization libraries, decomposition techniques (Column Generation, Benders), HPC/cloud deployment of systems, and quantum/quantum-inspired optimization.

Key Skills

Algorithm Design, Mathematical Modeling/DSL, High-Performance Scientific Computing, Performance Optimization, Solver Architecture.