Chorus Upgrade Shifts Machine Learning Emphasis
The latest version of Alpine Data’s analytics platform seeks to combine data with machine learning for business users by shifting the focus from algorithms while adding human collaboration and governance capabilities to machine learning projects.
San Francisco-based Alpine Data said Wednesday (May 25) its Chorus 6.0 analytics platform targets a wider swathe of the business analytics spectrum by accelerating “the delivery of data into action and to create a clear, repeatable process for continuous business improvement.”
Among the reasons that heavy enterprise investment in big data projects often does not pan out, the company argues, are the often-fruitless search for the “right” algorithm. Another pain point is enterprise struggles over data infrastructures and governance. Data governance refers to the overall management of data access rights and other security considerations.
“To generate far greater value from big data, organizations need to view data science and machine learning not as a project, but as integrated into the fabric of how they operate,” the company argued in announcing the latest version of the Chorus platform.
With the stated goal of helping organizations manage a “repeatable process that connects machine learning to business behavior,” Alpine said Chorus 6.0 adds Python integration with Jupyter Notebooks. Jupyter Notebook is a web application that allows users to create and share documents that contain live code, equations, visualizations and explanatory text.
That capability provides a faster, more flexible exploratory data analysis environment, the company asserted. The latest version of Chorus also aims to make it easier to read data from Hadoop clusters and databases into the Python environment. In addition, it makes Python scripts visible so they can be managed in Chorus. That, the company said, makes it easier to share and audit work.
The latest version also includes APIs for pulling data from configured relational databases and Hadoop clusters. Notebooks are executed, persisted and versioned on the server-side so data teams can collaborate in Python.
Chorus 6.0 also integrates data wrangling software developer Trifacta’s data transformation and cleansing capability, allowing analysts and data scientists to, among other things, discover, structure, clean and publish data for analysis. Trifacta’s proprietary software uses machine learning algorithms and a graphical user interface to accelerate the task of cleansing and preparing raw data for analyses.
The upshot, according to Steven Hillion, Alpine Data’s chief product officer, is that the upgraded analytics platform can help enterprise teams “organize and apply rigor to the data science process all the way from defining the business problem, through to an operational solution.”
Alpine Data said customers such as Morgan Stanley’s (NYSE: MS) wealth management team and the French advertising and public relations company Havas (EPA: HAV) currently use its Chorus analytics platform. Other users include medical benefits management specialist eviCore.