Case Study Deskcasestudydesk.com
Financial TechnologySourced

Case Study: Clear Street uses Metaplane to prevent $100M+ in data quality issues

Clear Street Case StudySourced & dated by Case Study Desk
Key facts · TL;DR
Company
Clear Street
Industry
Financial Technology
Challenge
On a prime brokerage platform, a listed price off by a penny can cascade into significant trading errors across hundreds of users.
Headline result
Clear Street deployed 300+ Metaplane monitors with 100% coverage of its dbt jobs, using data observability to prevent $100M+ worth of data quality issues on its prime brokerage platform

Key results

$100M+
Data quality issues prevented
prime brokerage platform
300+
Data quality monitors deployed
across critical tables
100%
Coverage of dbt jobs

The challenge

Clear Street runs a cloud-based modern data stack on Snowflake and Sigma that powers decision-making for a prime brokerage platform moving large volumes of money daily. With hundreds of attributes per financial instrument, a single incorrect value can change downstream valuations dramatically and, multiplied across users and transactions, produce significant losses. Although there were no pressing incidents, the team knew the risk of data quality issues would grow as the stack scaled.

The solution

Clear Street evaluated open-source and paid options before choosing Metaplane for its data quality monitors, integrations, and column-level lineage. The team integrated Snowflake, Sigma, dbt, and Fivetran, and deployed over 300 monitors tracking freshness, row count, cardinality, numeric distributions, and string formatting on critical objects such as ticker symbols.

Metaplane is the data quality x-ray on our data stack.

DW
David Wasserman
Senior Data Architect, Clear Street

The results, in context

Metaplane is positioned as the data quality layer that helps Clear Street prevent $100M+ worth of data quality issues on its platform. The team deployed more than 300 data quality monitors with 100% coverage of its dbt jobs, and used row-count monitoring to catch a client generating an anomalously high number of records, prompting a proactive customer conversation. Column-level lineage also supported a legacy-database deprecation by mapping downstream dependencies.

Products used

Metaplane Metaplane