Michael Z

At Plaid, I build transaction analytics platforms across AWS S3, Databricks, PySpark, Snowflake, dbt, and Airflow, reducing Product and Finance data availability from 4–6 hours to 60–90 minutes. I also develop replayable, idempotent transaction pipelines and historical reprocessing workflows that improve recovery, reporting reliability, and Risk investigation delivery.

Previously at Marqeta, I automated payment lifecycle reconciliation across authorization, clearing, and settlement data, reduced Finance report readiness from six to four hours, and enabled same-day month-end reporting. I built real-time payments analytics and secure customer-scoped reporting APIs using Kinesis, Spark Streaming, Snowflake, Lambda, and RBAC.

My earlier work at Etsy included marketplace analytics, seller reporting, Kafka event ingestion, experimentation datasets, and a transition from Hadoop and Oozie toward Airflow, GCS, and Dataproc. I enjoy building governed data models, improving data freshness, and mentoring engineers on reusable data engineering practices.