This is a fully remote position, open to applicants in Brazil.

📋 Description

• A 12-month initiative aimed at developing and assessing methods for optimizing cash denomination provisioning per withdrawal at ATM machines.

• Formalization of the issue, including variables, physical constraints, and the objective function, along with the creation of a simulation environment utilizing withdrawal history and Monte Carlo resampling.

• Development and evaluation of three distinct approaches: a periodic offline policy, a predict-then-optimize method combined with reinforcement learning, and a short-horizon online MIP with a fallback option.

• Coordination at the battery level, which encompasses a collection of terminals, to synchronize depletion rates and minimize overall restocking needs.

• Assessment against baseline models (largest-denomination, proportional, threshold) using metrics related to restocking events, the interval between restocks, service efficiency, latency, and traceability.

⛳️ Requirements

• PhD or Master’s degree.

• Currently pursuing or have completed studies in Computer Engineering, Computer Science, Electronic Engineering, Applied Mathematics, or Statistics.

• Experience with solvers such as CPLEX, Gurobi, CBC, HiGHS, etc.

• Proficiency in mathematical modeling techniques through Operations Research, including linear programming, mixed-integer programming, and heuristics.

• Programming skills in Python.