Herivelton Andreassa

I build data platforms, pipelines, and integrations that turn complex information into reliable data for business operations, analytics, and AI applications.

At Embroker, my work spans cloud data engineering and AI-powered automation. I’ve led pipeline migration to dbt, built staging and datastore layers in Snowflake, and developed ingestion workflows using Airflow, Stitch, and Matillion. I also build company enrichment pipelines and connect data services with Salesforce and internal platforms through MuleSoft and event-driven messaging.

Alongside this work, I develop pipelines that extract structured data from insurance documents and broker emails. These systems support application prefilling, quote generation, and document validation, combining LLMs and OCR with model fallbacks, monitoring, and automated reprocessing.

Previously, as Lead Data Engineer at BEES / AB InBev, I led the development of cloud data products supporting operations across 20+ countries using Azure, Databricks, dbt, and Snowflake. I also helped hire and onboard 10+ data professionals.

My core stack includes Python, SQL, AWS, Azure, Snowflake, Databricks, dbt, and Airflow. I’m particularly interested in data architecture, pipeline reliability, cloud cost optimization, and building the data foundations that make AI useful in production.