Open Source

Bloomberg sits at the intersection of high-availability, low-latency and large-scale computing, which leads us to take a very pragmatic approach to using open source software to address complex problems. We choose open source for new infrastructure needs even when it may only meet 80% of our requirements.

Because of the unique technical constraints we face in finance, we prefer tweaking and extending high-quality, web scale open source solutions to address our needs rather than using proprietary or source-available solutions.

Our engineers are active members of the open source community, contributing their skills and passion to a wide variety of initiatives. Hundreds of our employees have contributed code, documentation, tests or other significant enhancements and improvements to projects for the benefit of the tech community.

In some areas especially relevant to our compute and data science infrastructure, Bloomberg engineers have earned leadership positions as committers, maintainers, technical steering committee (TSC) and project management community (PMC) members. They also sit on foundation boards and standards committees, including those directing the future of the C++, Python and JavaScript programming languages.

We are committed to sustaining and broadening the scope of innovation through direct collaboration with open source communities. In academia, Bloomberg funds research that produces new or enhanced libraries, as well as teams that directly create and extend projects relevant to our products, including Project Jupyter and JupyterLab.

We host events, sponsor projects and provide access to the knowledge and experience of our engineers. Since 2013, we have hosted Open Source Day/Weekend events that bring together our employees, community members and students to sprint on open issues in an open source project.

Projects that have benefited include Git, Clang/LLVM, Eclipse, Python, Perl, pandas, NumPy, SciPy and Matplotlib. We also fund improvements to open source projects that are important to our engineers, such as the development of CSS Grid Layout in Chromium, the addition of BigInt and other features in WebAssembly, and improvements in GNU Autoconf and version 3 of Psycopg.

Each year, we proudly provide financial support to a variety of technology community organizations, in addition to sponsoring and speaking at dozens of relevant conferences.

In an effort to enable direct employee engagement to help sustain the open source projects critical to our work, Bloomberg’s OSPO has partnered with the company’s Corporate Philanthropy team to launch the Bloomberg FOSS Contributor Fund.

Open source tools created at Bloomberg to solve problems with cloud infrastructure, information retrieval, data science, natural language processing, mobile application development and more have been published on GitHub for others to use in solving real-world problems. Examples include PowerfulSeal, ipydatagrid, Comdb2, Goldpinger, xcdiff, and amqpprox.

OUR ORGANIZATION

Several open source projects born at Bloomberg have been spun out and are now supported by communities with their own governance:

• The Solr Operator, developed by Bloomberg engineers to manage Apache Solr Clouds within Kubernetes, was donated to the Apache Solr project.
• Our bqplot interactive plotting and charting library for Jupyter notebooks is now a NumFOCUS Affiliated Project.
• KServe, a simple, yet powerful, Kubernetes Custom Resource for deploying machine learning (ML) models at scale, was co-developed and led by Bloomberg engineers prior to the community contributing it to the LF AI & Data Foundation, where it is now a full-fledged Incubation Project.