article thumbnail

Accelerating AI at scale without sacrificing security

CIO Business Intelligence

Similarly, in 2017 Equifax suffered a data breach that exposed the personal data of nearly 150 million people. Examples include the 2008 breach of Société Générale , one of France’s largest banks, when an employee bypassed internal controls to make unauthorized trades, leading to billions of dollars lost.

article thumbnail

How companies are building sustainable AI and ML initiatives

O'Reilly on Data

In 2017, we published “ How Companies Are Putting AI to Work Through Deep Learning ,” a report based on a survey we ran aiming to help leaders better understand how organizations are applying AI through deep learning. Data scientists and data engineers are in demand.

Insiders

Sign Up for our Newsletter

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.

article thumbnail

What are model governance and model operations?

O'Reilly on Data

Metadata and artifacts needed for a full audit trail. An overview from a 2017 paper from Google lets us gauge how much tooling is still needed for model operations and testing. At the moment, few (if any) teams have checklists as extensive as the one detailed in the 2017 paper from Google.

Modeling 257
article thumbnail

Deep automation in machine learning

O'Reilly on Data

Metadata analysis makes it possible to build data catalogs, which in turn allow humans to discover data that’s relevant to their projects. We now need to build the tools for software+data: tools to track data provenance and lineage, tools to build catalogs from metadata, tools to do fundamental operations like ETL.

article thumbnail

Gartner Magic Quadrant for Metadata Management Includes Alation

Alation

Gartner predicts that “By 2020, 50% of information governance initiatives will be enacted with policies based on metadata alone.”. Magic Quadrant for Metadata Management Solutions , Guido de Simoni and Roxane Edjlali, August 10, 2017. Metadata management no longer refers to a static technical repository.

article thumbnail

Themes and Conferences per Pacoid, Episode 11

Domino Data Lab

In other words, using metadata about data science work to generate code. One of the longer-term trends that we’re seeing with Airflow , and so on, is to externalize graph-based metadata and leverage it beyond the lifecycle of a single SQL query, making our workflows smarter and more robust. BTW, videos for Rev2 are up: [link].

Metadata 105
article thumbnail

5 Hardware Accelerators Every Data Scientist Should Leverage

Smart Data Collective

Although SageMaker has become a popular hardware accelerator since it was launched in 2017, there are plenty of other overlooked hardware accelerators on the market. It allows data scientists to log, store, share, compare and search important metadata that is used to build models for data science applications. Neptune.ai. Neptune.AI