Juan Garcia Sanchez
Juan Garcia Sanchez
About
Detail
VP of Engineering, Platform
Ruse, Bulgaria
I am a Senior Data Scientist and Lead Instructor specialized in building end-to-end machine learning systems, engineering big data architectures, and translating complex data telemetry into strategic engineering solutions. My career spans managing massive data pipelines for Fortune 500 enterprise clients, scaling tech stacks for high-growth startups, and mentoring the next generation of data professionals.
Currently, as a Senior Data Scientist at Zeta Global, I architect end-to-end machine learning pipelines and optimize Big Data workflows for premier US enterprise clients. My focus is on technical leadership, workflow automation, and high-throughput model deployment. By refactoring code and building custom, semi-automated Python and R libraries, I successfully engineered optimizations that slashed our team's modeling pipeline runtimes from 40 hours down to just 8 hours.
Previously, at GoTrendier, I supported a circular-economy marketplace by designing cloud data warehouses on AWS Redshift, orchestrating ETL/ELT pipelines with Apache Airflow, and deploying custom machine learning models. This includes Time-Series anomaly systems, Click-Through Rate (CTR) algorithms, and Natural Language Processing (NLP) solutions for unstructured text.
Driven by a passion for the data community, I have also led and designed technical training cohorts globally across Ironhack, Develhope, and Masterschool, instructing students in advanced SQL, version control, data storytelling, and production-ready architectures.
Currently, as a Senior Data Scientist at Zeta Global, I architect end-to-end machine learning pipelines and optimize Big Data workflows for premier US enterprise clients. My focus is on technical leadership, workflow automation, and high-throughput model deployment. By refactoring code and building custom, semi-automated Python and R libraries, I successfully engineered optimizations that slashed our team's modeling pipeline runtimes from 40 hours down to just 8 hours.
Previously, at GoTrendier, I supported a circular-economy marketplace by designing cloud data warehouses on AWS Redshift, orchestrating ETL/ELT pipelines with Apache Airflow, and deploying custom machine learning models. This includes Time-Series anomaly systems, Click-Through Rate (CTR) algorithms, and Natural Language Processing (NLP) solutions for unstructured text.
Driven by a passion for the data community, I have also led and designed technical training cohorts globally across Ironhack, Develhope, and Masterschool, instructing students in advanced SQL, version control, data storytelling, and production-ready architectures.
