Back to work
Supply chain AI data platform
productionNoodle.ai, 2019–2022
End-to-end pipeline for a first-in-class supply chain AI product, processing TB-scale customer data with 99.9% uptime.
SnowflakeSparkAirflowPythonDocker
Problem
Each new customer meant re-implementing the pipeline by hand. Implementation time was the bottleneck on growth, and the team was two engineers.
Approach
- Separated business logic from code with configurable, customer-agnostic data models deployable across accounts.
- Built a reusable Airflow framework so engineers could write new pipelines quickly and consistently.
- Grew and led the data engineering team from 2 to 8 engineers.
Impact
- Customer implementation time cut by 50%.
- Pipeline development accelerated by 60% across teams.
- 100% team retention; three engineers promoted to senior roles.