Machine learning data science platform drops by SnapLogic
|Christian Hargrave in Artificial Intelligence Tuesday, November 20, 2018|
New SnapLogic Data Science solution improves data engineering and data science results by accelerating the development and deployment of machine learning models.
SnapLogic announced SnapLogic Data Science, a new self-service solution to accelerate the development and deployment of machine learning with minimal coding. Through SnapLogic’s drag-and-drop interface, data engineers, data scientists, and IT/DevOps teams can use SnapLogic Data Science to manage and control the entire machine learning lifecycle - including data acquisition, data exploration and preparation, model training and validation, and model deployment - all from within the SnapLogic integration platform. SnapLogic Data Science breaks down traditional barriers that can undermine machine learning initiatives by providing a common platform for machine learning visibility and collaboration across teams including data engineering, data science, IT, DevOps, and development.
According to their recent research with Vanson Bourne, 68% of IT decision-makers consider artificial intelligence and machine learning as vital to accelerating their transformation projects. At the same time, McKinsey Global Institute predicts that the U.S. alone will be short 250,000 data scientists by 2024. Machine learning initiatives are hampered by limited access to data science talent as well as a lack of automated data access to fuel model building. By bridging the data science skills gap and automating the machine learning lifecycle, SnapLogic Data Science makes end-to-end machine learning accessible to enterprises of all sizes for the first time.
“Every enterprise in every industry will need to employ AI and machine learning in order to keep pace with today’s most progressive businesses. However, most companies fall flat in actualizing machine learning because they don’t have the talent or financial resources to make the most of their data,” said Greg Benson, Chief Scientist at SnapLogic. “With SnapLogic Data Science, we’re enabling our customers to overcome the common barriers associated with putting machine learning into practice by arming them with a full stack of self-service tools to be faster, more agile, more data-driven. Just as we enabled self-service application and data integration for IT and citizen integrators, we are extending these self-service capabilities to data engineers and data science teams who need to build and deploy machine learning models faster and easier.”
“451 Research believes the operationalization of data science projects involving machine learning and other artificial intelligence technologies is set to be a significant aspect of the next wave of developments in the data management space,” said Matt Aslett, Research Vice President, Data, AI and Analytics, 451 Research. “As adoption of data science expands and matures we expect to see enterprises looking for products and services that simplify and support the complete machine learning lifecycle from development, through training and testing to deployment.”
Putting Machine Learning into Practice
Machine learning is an integration problem. Until now, no other solution on the market has solved the problem of data integration and machine learning operationalization in a single platform. SnapLogic Data Science provides enterprises with self-service tools to rapid.ly build and deploy machine learning models from beginning to end. By democratizing end-to-end data science, organizations can achieve higher productivity through accelerated machine learning development following a visual, drag-and-drop interface. Enterprises also achieve a lower total cost of ownership as they are now able to decrease their production deployment time from days to hours, natively within the SnapLogic integration platform.
With SnapLogic Data Science, organizations can:
- Access and assemble the data for use by machine learning models
- Perform preparatory operations on data sets such as data type transformation, data cleanup, sampling, shuffling, and scaling
- Create, cross-validate, and deploy machine learning models. Additionally, users can also execute Python scripts remotely to leverage libraries such as TensorFlow and Keras
- Perform analytic operations such as data profiling and data type inspection
- Eliminate redundancy in data preparation introduced through the disconnect between data engineering, data science, and IT/DevOps teams
- Operationalize machine learning models built using SnapLogic’s new Machine Learning Snaps or Jupyter Notebooks with the Native Python Snap
- Deploy machine learning pipelines as Ultra Tasks to receive and reply to machine learning API requests
- Provide continuous model building to adapt to new data and activity