Luke Smith

Data Engineer

I build data systems — pipelines, models, and tooling — and the physical machines they measure and predict. Mechanical engineer turned data engineer, working mostly in data and ML infrastructure.

Selected work
Clickstream Sessionization01 / 05

GA4 ingest · data normalization · session metrics

Product Metadata Enrichment02 / 05

hybrid discriminative + generative AI · hierarchical modeling · human-in-loop review

Synthetic Control Experiments03 / 05

control group selection · significance testing · internal tooling

Micro-pump Priming Chamber04 / 05

resistive level sensing · micron-scale intake filtering · mixing

Gasket Install System for End-of-Arm Tool Changer05 / 05

fixturing · pneumatic tooling · time study

Experience
Data EngineerJul 2021 — Present
Valtech · Remote

GA4 clickstream fact + feature tables; international price-experimentation infrastructure (PySpark, Databricks); generative product-metadata enrichment.

Founding ConsultantJul 2020 — Jul 2021
Lassometrics · NC

Serverless AWS data pipelines (Step Functions, Lambda, S3, Athena); report automation; web-scraping ingestion.

R&D EngineerJul 2018 — Jul 2020
Coca-Cola · GA

Manufacturing data dashboard (Python, AWS); flowmeter filter optimization; patent-pending injection-molded components.

Manufacturing EngineerMay 2016 — Aug 2016
ATI Industrial Automation · NC

Increased bushing-install tool throughput 15% via 3D-printed and purchased tooling.

Stack
Python · SQL · Spark · Airflow · Databricks · PyTorch · Hugging Face · MLflow · Docker · Terraform · AWS
Contact
contact@lukesmith.engineer