Apache Beam at Postmates

[Dmitry Demeshchuk <dm...@postmates.com>]

Hi folks,

Hope this kind of email is appropriate for this mailing list.

Postmates deals with the real world and dense urban areas in a unique way,
we use events to dynamically respond to changes in the world. Postmates is
looking for a full-time engineer who would focus on changing the way
Postmates generates, processes, and reasons about events, primarily helping
us adopting Apache Beam faster and more efficiently.

We have two major use cases for Beam:

1. Good old ETL batch processing jobs.
2. Consuming real time events coming from across Postmates systems and
processing them in various ways. Some usage examples could be:
    * Fraud detection
    * Courier dispatch mechanisms
    * React to changes in a city
    * More up-to-date analytics for the Marketing team

We use the Beam Python SDK (although the Java SDK is not completely off the
table) and use Google Dataflow as a runner.

We are looking for a person (or people) who would be willing to help us
drive it. Here are some things you'll likely get to work on:

* Creating more sources and sinks for Beam (most of our stack currently
runs on AWS).
* Helping to improve the Beam Python SDK by contributing to it.
* Adapting Beam for easier usage within the company (simplifying some
developer interfaces, adding commonly used PTransforms, etc).
* Creating and driving the user stories for data analysts and data
engineers (you'll get to collaborate with both).
* Helping unify our data pipelining: reporting, transporting, storing and
reading.

Bottom line is: you will be getting paid for being a Beam user and
contributor!

If you are interested, drop an email to "dmitry at postmates.com" or "emad
at postmates.com", we'd be happy to tell you more!

-- 
Best regards,
Dmitry Demeshchuk.