Research Intern — AI and Quantum Algorithms for Optimization 2027

At IBM Research, we build what's next in computing. Our teams work at the intersection of quantum information science, artificial intelligence, and high-performance computing — advancing quantum-centric supercomputing, where quantum processors and classical HPC resources work together to solve problems that neither can address alone. As a research intern, you will join a team of scientists and engineers working on the algorithmic foundations of this vision: designing, analyzing, and benchmarking quantum and hybrid quantum-classical algorithms for optimization and scientific computing, and exploring how AI and agentic systems can accelerate algorithm discovery and workflow orchestration. You will have access to IBM's fleet of utility-scale quantum computers, the Qiskit software stack, and the classical computing infrastructure that surrounds them. This is a hands-on research role. Interns are expected to contribute original technical work, and strong outcomes frequently lead to publications at leading venues, open-source contributions, and continued collaboration with IBM Research. You will work with IBM Research scientists on the design and analysis of algorithms for quantum and quantum-centric supercomputing, with a focus on optimization. Specifically, you will: Design, implement, and analyze quantum, classical, and hybrid quantum-classical algorithms for optimization problems, including model-based (e.g. MILP, conic, nonlinear) and data-driven formulations. Develop and benchmark quantum optimization approaches — variational and non-variational methods, quantum-enhanced heuristics, and circuit-cutting or sampling-based hybrid workflows — on IBM quantum hardware and simulators. Contribute to quantum-centric supercomputing workflows that partition problems across QPUs and classical HPC resources, and characterize where quantum resources provide advantage. Establish rigorous performance baselines against state-of-the-art classical solvers, and carry out complexity, scaling, and resource-estimation analyses. Explore the use of AI and agentic systems for algorithm design, hyperparameter and ansatz search, code generation, and automated experiment orchestration. Implement research prototypes in Python (Qiskit and the broader scientific Python ecosystem), with clean, reproducible, and well-documented code. Present results in team meetings, contribute to technical reports, papers, and patent disclosures, and where appropriate contribute to open-source projects. Quantum computing (required). Solid working knowledge of quantum information and quantum algorithms — circuit model, Hamiltonian simulation, variational and sampling-based algorithms, noise and error mitigation — with practical experience implementing and running circuits (e.g. Qiskit). Strong mathematical and algorithmic foundations: linear algebra, probability, discrete mathematics and combinatorics, algorithm design and analysis, and computational complexity. Programming proficiency in Python, including scientific and numerical libraries (NumPy, SciPy), plus the software discipline to produce reproducible experiments and readable, version-controlled code. Demonstrated research ability: framing a problem precisely, designing and running rigorous experiments, and communicating results clearly in writing and in talks. Depth in one or more of the following, in addition to quantum: model-based optimization (linear/integer/convex/nonlinear programming, metaheuristics); data-driven optimization and machine learning (including learning-to-optimize and surrogate models); quantum optimization; quantum-centric supercomputing and hybrid quantum-classical algorithm development; agentic AI and LLM-based systems; theory of computation. Publications or preprints in quantum computing, optimization, theoretical computer science, or machine learning venues. Hands-on experience running experiments on real quantum hardware, including transpilation, error mitigation and suppression, and interpreting hardware noise. Experience with commercial or open-source optimization solvers (CPLEX, Gurobi, MOSEK, SCIP, OR-Tools) and modeling frameworks (JuMP, Pyomo, CVXPY, DOcplex). Experience with HPC environments: distributed and parallel computing, GPU acceleration, job schedulers, and tensor-network or large-scale simulation tooling. Experience building agentic AI systems — tool use, multi-agent orchestration, retrieval, and evaluation of LLM-driven workflows — particularly applied to scientific or mathematical problem solving. Familiarity with quantum error correction, fault-tolerant algorithm design, or resource estimation. Open-source contributions, especially to Qiskit or related quantum or optimization libraries. Additional programming experience in C/C++, Julia, or Rust. Fluency in written and spoken English.