Ravit Pichayavet

About

Hi! I'm Ravit Pichayavet. I'm an Applied Scientist at Amazon, in Amazon Fulfillment Technology (AFT) Science, working on optimization for supply chain and logistics. I recently finished my Ph.D. at Georgia Tech's ISyE, advised by Professor Alan Erera, and defended in July 2026 with the degree conferred in December 2026.

My work is about high-velocity logistics: the systems behind fast delivery, where speed, cost, and limited capacity all pull against each other. That spans the last mile, routing vehicles through dense cities under delivery-time guarantees, and the middle mile, planning freight flows across national networks and buying truckload capacity ahead of uncertain demand.

My recent CV (last updated November 2026).

Industry Experience

Applied Scientist

Amazon Fulfillment Technology (AFT) Science · Amazon, Bellevue, WA · Aug 2026 – Present

Applied Scientist Intern

Middle Mile Science, LTL Team · Amazon, Bellevue, WA · May 2025 – Aug 2025

Designed and implemented a routing optimization module (e.g., Mixed-Integer Programming, Column Generation, Routing heuristics) to improve the efficiency of LTL freight movement in Amazon's middle-mile network.

Teaching Experience

Graduate Student Instructor

ISyE 3103: Supply Chain Modeling: Logistics · Georgia Tech · Spring 2026

Served as the instructor for the undergraduate course ISyE 3103 Supply Chain Modeling: Logistics, delivering lectures on inventory management, freight transportation, and supply chain network design.

Research

Rate-based Vehicle Routing Problem for Delivery in Densely Populated Urban Areas

A branch-and-price method for last-mile routing when orders arrive continuously and every delivery carries a lead-time guarantee.

Improvement Heuristics for Extremely Large-Scale LTL Flow and Load Planning

Local search heuristics that improve flow and load plans for parcel consolidation networks at U.S. national scale.

Advance Reservation of 3P-Freight Capacity under Demand and Capacity Uncertainty

A two-stage stochastic model for reserving truckload capacity in advance when both shipment volume and carrier availability are uncertain, with an exact solution method and a predict-then-optimize deployment.

Thesis: Data-Driven Optimization of High-Velocity Logistics Systems: Addressing Challenges from Last-mile Delivery to Middle-mile Freight Management

Successfully defended July 2026 · Degree conferred December 2026
Major in Supply Chain Engineering, minor in Computation and Machine Learning

The acknowledgments are my favorite part. → Read them here