Follow Datanami:
October 19, 2020

Goldman Releases Fintech Data Modeling Tool

(Bruce Rolff/Shutterstock)

Goldman Sachs is launching an open source data modeling platform based on its flagship data management and governance platform.

The investment banking giant (NYSE: GS) said the open-source version of the platform dubbed Legend includes source code for five modules within its internally developed version. The modules are available as open source through the Fintech Open Source Foundation (FINOS).

The Linux Foundation acquired the operating assets of FINOS in April. The group was formed in 2014 to promote adoption of open source software and standards in the financial services sector.

The community release follows a six-month pilot program during which investment banks shared Legend on public cloud infrastructure hosted by FINOS. The pilot included collaboration among banks to develop a prototype system for data modeling and standardization. In one example, collaborators built extensions to the Common Domain Model, a data model used to standardize the trading of derivatives.

Legend “is generally available for industry participants to use and build models collaboratively,” FINOS said Monday (Oct. 19). Goldman Sachs is a platinum member of the organization, and has aggressively pursued fintech standardization efforts.

In development since 2013, Legend would help “automate some of the most difficult data governance challenges, as well as provide self-service tools to democratize data and analytics,” said Atte Lahtiranta, Goldman Sachs’ CTO. It would also provide “greater standardization and efficiency across the entire financial services ecosystem.”

The investment bank’s internal data management strategy focuses on tracking the full cycle of transactions from trade confirmation to settlements while meeting regulatory requirements. Internally, the platform and language were known as “Alloy” and “PURE,” said Pierre De Belen, head architect of the Legend platform.

Goldman Sachs joins a growing list of fintech developers releasing internal data management platforms to the open source community. Morgan Stanley released its Morphir project to FINOS in September, and said this week it expects to achieve synergies with Legend.

FINOS promotes Legend as a tool for maintaining consistent data across different applications and databases. Elements of complex financial instruments are often scattered across hundreds of systems within an investment bank’s infrastructure. Platforms like Legend provide “advanced data lineage tooling,” said Rob Underwood, chief development officer at FINOS and leader of the open-sourcing effort.

Legend will “help drive data consistency and improve interoperability [that] will compound as it’s adopted by more institutions, both sell-side and buy-side,” added Underwood.

Legend source code is here. The five modules are Legend Studio; Legend Engine, Legend SDLC and Legend Shared, along with the underlying PURE data modeling language.

Recent items:

For ‘Open Banking,’ Data is a Person

Staying on Top of ML Model and Data Drift

Datanami