Kévin Hautmann

Worked within Amazon’s Central Operations team, the control tower overseeing last-mile delivery performance across Europe.

My role focused on improving fleet planning efficiency, route execution, and operational decision-making through data engineering, process automation, BI dashboards, and analytics solutions, while collaborating closely with technical and operational stakeholders across multiple teams.

Key Achievements :

- Designed and implemented new logic in an AWS Lambda workflow to optimize delivery van utilization through route rebalancing recommendations across 350+ delivery stations, contributing to $50M+ annual savings.

- Built a QuickSight operational dashboard used during peak events to identify where route cancellations should be limited, improving Delivery Service Partner engagement and delivery reliability.

- Developed a Python / Tkinter internal tool automating route cancellation parameter updates across EU sites, replacing manual JSON edits and saving 1.5 hours per week for operations managers while reducing execution errors.

- Authored an 11-page SOP enabling a team of 10+ managers to independently manage route parameter changes, reducing dependency on technical teams and freeing bandwidth for higher-value initiatives.

Technical Environment: AWS (S3, Lambda, Redshift, Glue, QuickSight, CloudWatch), Python, SQL, DBT, ETL / ELT, Linux, Git, BI/Observability, Data Automation, Operational Analytics

Joined Mobivatar during the early development phase of their video game Foodogeki, at a stage where business data was fragmented across JSON files and the existing database lacked the structure required for future growth.

My role focused on building a reliable and scalable PostgreSQL foundation through database architecture, data modeling, automation, cost optimization, and production administration, while working closely with product, development, and business stakeholders.

Key Achievements :

- Redesigned and normalized the PostgreSQL production database, reducing storage costs by 40% while improving query performance and maintainability.

- Designed and documented 140 tables, 26 triggers, and multiple views, creating a robust data model aligned with gameplay and business requirements.

- Implemented automated schema versioning, backup & recovery processes, and role-based access controls, improving reliability and reducing manual deployment risks.

- Deployed production and development databases on GCP Cloud SQL, and reduced development environment costs by 65% by automating instance shutdown during idle periods.

Technical Ecosystem : PostgreSQL, Query Optimization, Data Modeling, Liquibase, Python, Bash, Linux, GCP Cloud SQL, Backup & Recovery, RBAC, Git

Worked within Amazon’s Vendor Flex Analytics team, supporting a program that integrates vendor warehouses into Amazon’s logistics network to accelerate deliveries and improve inventory efficiency across Europe.

My role focused on improving operational visibility and decision-making through data engineering, BI dashboards, ETL pipelines, and data quality solutions, while collaborating closely with business stakeholders, operations teams, and analysts across multiple countries.

- Built and maintained a QuickSight dashboard tracking 15+ operational and retail KPIs, used by 30+ stakeholders to drive weekly performance reviews across Vendor Flex sites in 6+ European countries.

- Developed and optimized 10+ ETL pipelines using Redshift SQL, processing datasets up to 7TB into Data Lake and Data Warehouse environments.

- Improved reporting reliability by implementing SQL-based data quality checks and validation layers, increasing stakeholder trust in operational metrics.

- Consolidated multiple reporting sources into streamlined BI solutions, reducing dashboard fragmentation and simplifying access to business insights.

Technical Environment : AWS (S3, Redshift, QuickSight), SQL, Python, DBT, ETL / ELT, Data Warehouse, Data Lake, KPI Reporting, BI, Data Quality