Case Study: Autodesk achieves unified data architecture and eliminates pipeline maintenance with Fivetran

Autodesk builds a culture of self-service data for 13,000+ employees

Autodesk Construction Services faced a fragmented data architecture after acquiring products like BuildingConnected and BIM 360, leaving the team juggling disparate applications, databases and analytics tools. The complexity caused heavy pipeline maintenance (about 3–5% of analyst time), thousands of lines of Python to unpack JSON, concurrency and storage limits, and long runtimes; Autodesk addressed this by adopting a modern data stack with Fivetran for pipelines, Snowflake for warehousing, and dbt for transformations.

Fivetran’s automated connectors, schema migrations, alerts and automatic backfill—paired with Snowflake’s elastic compute and dbt’s repeatable transformations—removed the need for manual backfills and over 3,000 lines of Python, cut BuildingConnected transformation runtimes by 68%, reduced pipeline maintenance to under 1% of analyst time, and turned BIM 360 extractions and transformations into minutes instead of hours while eliminating daily monitoring.

Autodesk

Evin Anderson

Data Engineering Manager