I used reinforcement learning to help the robot make better decisions about what to do and when to do it in the wonderfully chaotic environment of a fast-food kitchen. I also made sure data got from sensors to dashboards reliably and on time so teams could actually use it. (More technical details below.)

Reinforcement learning scheduling algorithms

• extended the simulation’s state representation

• tuned hyperparameters for stable, higher-performing training

• built visualization tooling to make policy behavior interpretable and faster to debug

Data platform

• designed and implemented AWS pipelines with Terraform

• established stable multi-environment production practices and managed releases

• delivered reporting/observability (recurring customer reports, dashboards, and alerts) for real-time fleet health and long-term trends.

Reinforcement learning for scheduling algorithms

● Developed and implemented efficient slotting algorithms to optimize warehouse productivity