Senior Software Engineer - Ticker Plant, Frankfurt

The Ticker Plant system is the beating heart of Bloomberg’s market data infrastructure and is responsible for the storage and distribution of all Market Data within Bloomberg. Market Data is information -- trades, quotes, and other pricing data -- gathered from around the world on every possible kind of financial instrument, from stocks and bonds to currencies and commodities, and other trading venues across the globe. In the aggregate, this amounts to several hundreds of billions of events across 100 million instruments.  Read our blog to know more!

Our Teams

Ticker Plant is organized into the following functional areas.

Market Data Distribution handles the movement of real-time market data in and out of the Ticker Plant system.  This group owns and develops infrastructure that carries hundreds of billions of events originating from across thousands of machines spanning hundreds of clusters with low latency.  In addition to the distribution infrastructure, this group provides sharding infrastructure that enables horizontal scalability of the Ticker Plant.

Database Infrastructure develops a high performance time-series database that supports low latency data storage and retrieval for high-volume applications, and stores all Market Data at Bloomberg. The storage engine hosts several petabytes of data, across hundreds of millions of instruments.  The query engine handles ~100 billion queries daily ranging from interactive queries to batch queries at very high throughputs of up to 100K QPS. 

Data Content area focuses on APIs, systems and services which transform, enrich, store and publish these market data events, relying on externally provided middleware and persistence technology solutions.

We are looking for motivated engineers in Database Infrastructure and Market Data Distribution areas to help build the next generation Ticker Plant! 

Our Technology

Being the core of the market data infrastructure, the Ticker Plant system by necessity is scalable, distributed, highly performant, and stable.  To be highly performant, most of the critical software is written in C++. Python is also extensively used to build auxiliary services around the critically important software and to enable automation.  In addition, we use a wide range of engineering strategies to solve problems ranging from using off the shelf software (like kafka) to implementing our own algorithms/solutions (e.g custom built database), as appropriate for the problem at hand. 

What we do

We have some unique challenges where we not only have to optimize for scale (billions of data points, hundreds of hosts) but also for very high throughput of updates and queries, all at the same time.  As market data volume increases each year, these challenges get harder to meet with the current infrastructure and so we are re-architecting and re-writing the core parts of our system to make the system simpler and easier to scale.  Some examples are:

• Consolidate existing data distribution mechanisms onto a fast, loss-less, resilient market data distribution infrastructure.

• Improve scalability of the Ticker Plant by separat