Last week’s biggest analytics-related news came not from the traditional data management vendors that usually dominate the agenda, but rather the web-scale crowd. Twitter Inc. led the  charge by publishing the code for Heron, a homegrown stream processing framework that its engineers use to glean high-level insights from tweets.

The engine is described as between two and five times more efficient than its predecessor, Storm, which the social media giant open-sourced as well a few years ago. Moreover, Twitter says that Heron is easier to maintain too thanks to a modularized design that makes it possible to quickly identify faulty analytic workflows in need of troubleshooting. These features should make the framework very appealing to the growing number of companies that seek to the fast-moving data coming off their connected devices.

However, the ability to ingest machine-generated transmissions while they’re still fresh is only one of the requisites for a stream processing project. Equally necessary is a machine learning component that can quickly separate the signal from the noise, functionality that Amazon Inc. is reportedly looking to provide for users of its public cloud. Word leaked last week that the vendor is developing a managed service designed to let customers build and deploy analytic algorithms on its infrastructure using open-source AI engines like Google TensorFlow.

If and when it launches, the offering will provide a valuable new way for Amazon to monetize companies’ fast-growing interest in artificial intelligence  and machine-generated transmissions. The revenue opportunity has also attracted the attention of fellow tech giants like Intel Corp., which joined the fray last week by acquiring a startup called Itseez Inc. that specializes in computing vision. It’s created a suite of algorithms for identifying traffic signs, potentially dangerous changes in driver behavior and other traffic threats. The software should enable the chip maker to establish a bigger presence in the nascent autonomous vehicle space, which is poised to become a major market for embedded processors. 

Image via Geralt