Kedro

2019: Announcement

At the beginning of June, 2019 the consulting company McKinsey issued the first open source software. The tool created on Python under the name Kedro intended for development of processes of pipeline processing of data became such product.

Software allows to structure an analysis code, bringing it into a uniform format and training him for commercial operation. Besides, Kedro helps to build modular pipelines for data processing and to separate them into versions.

McKinsey provided the first to software open source

The technology was created by programmers Nikolaos Tsaousis and Aris Valtazanos from QuantumBlack company which McKinsey purchased in 2015.

According to the CEO of QuantumBlack Jeremy Palmer, Kedro — this "a big step" for McKinsey as the company "continues to counterbalance income from own assets and a possibility of participation in community of developers".

In  a conversation with the Computerworld UK edition the head of department of development and promotion of the products QuantumBlack Michele Battellli noted that Kedro represents library of codes which can be used for creation of pipelines of data and machine learning — "in essence construction blocks of what we do in analytics or projects of machine learning".

McKinsey and other companies which decided to develop the Open Source projects with their help expect to involve developers who most often prefer to study fundamental technologies, but not the systems of certain producers. Thanks to support of third-party developers McKinsey in the future will be able to make Kedro a commercial product when understands that it is developed enough for sale. Besides, the company hopes that this platform will become the industry standard for the industrial code in the field of machine learning and data processing.^[1]

Notes