Warehouse Order Batching

As Senior Applied Scientist at Zalando, I developed and implemented warehouse order batching and picker routing algorithms that reduced item picking costs by approximately 5%, across 500 million warehouse items processed yearly.

The core insight: warehouse operations involve four interconnected decisions. Which orders to pick, which items to allocate, how to batch orders together, and how to route pickers. Traditional systems solve these sequentially, but solving them jointly with mixed-integer programming yields significant efficiency gains. We published a research paper on the problem formulation and released benchmark instances to the research community.

Beyond the batching work, I architected an algorithm development pipeline that reduced the time from prototyping new algorithms to production deployment from months to days. This infrastructure accelerated iteration across the logistics algorithms team. I also designed and implemented a service for inbound article sortation and maintained legacy warehouse planning systems.

I was promoted from Mid to Senior Applied Scientist after 1.5 years.