Curated for content, computing, and digital experience professionals

Day: September 28, 2022

AtScale adds enterprise AI capabilities to semantic layer platform 

AtScale, a provider of semantic layer solutions for modern business intelligence and data science teams, announced new product capabilities for organizations working to accelerate the deployment of enterprise artificial intelligence (AI). These new capabilities leverage AtScale’s position within the data stack with cloud data warehouse and lakehouse platforms including Google BigQuery, Microsoft Azure Synapse, Amazon Redshift, Snowflake, and Databricks. The AtScale Enterprise semantic layer platform now incorporates:

  • Semantic Predictions – Predictions generated by deployed AI/ML models can be written back to cloud data platforms through AtScale. These model-generated predictive statistics inherit semantic model intelligence, including dimensional consistency and discoverability. Predictions are immediately available for exploration by business users using BI tools (AtScale supports connectivity to Looker, PowerBI, Tableau, and Excel) and can be incorporated into augmented analytics resources.
  • Managed Features – AtScale creates a hub of centrally governed metrics and dimensional hierarchies that can be used to create a set of managed features for AI/ML models. AtScale managed features inherit semantic context, making them more discoverable and easier to work with. Managed features can now be served directly from AtScale, or through a feature store like FEAST, to train models in AutoML or other AI platforms.

Gilbane Advisor 9-28-22 — MLOps or no, Tabu search, practical NLP

This week we feature an article from Lak Lakshmanan.

Additional reading comes from Adam Langley, Fabio Chiusano, and Jing Huang, Mingang Fu, & Minghui Liu.

News comes from Apptek &, Fivetran, Cloudflare, and Stardog & Databricks.

If you’ve missed any of the past 57 issues you can see them here; those, as well as all earlier issues can be found here.

Opinion / Analysis

No, you don’t need MLOps

Well you might. But either way Lak Lakshmanan provides lots to think, or re-think, about. (14 min).

MLOps started from a straightforward problem statement — that the technical debt associated with ML models becomes intolerable if the models are not adjusted over time to account for changes in the environment. Since that 2015 observation, ML models and frameworks have been built that make it relatively easy to avoid the most glaring potholes in the way of the ML practitioner. However, in the past year or so, the MLOps buzzword has taken on a life of its own. At this point, most things sold as MLOps are overkill and unnecessary for most teams.

All Gilbane Advisor issues

More Reading

Content technology news

Stardog joins Databricks Partner Connect

The integration means organizations can add a unified semantic layer atop Databricks to accelerate data analytics for more data-informed decisions.

Fivetran introduces Metadata API

Customers can integrate with governance and observability tools to give data teams more control over who has access to what data.

Apptek and announce strategic partnership

To help companies augment intelligent automation by extending AI-based text analytics to audio content, even across multiple languages.

Cloudflare launches Data Localization Suite in Asia

To help businesses comply with data localization obligations by using Cloudflare to set rules and controls on data storage ad access.

All content technology news

The Gilbane Advisor is curated by Frank Gilbane for content technology, computing, and digital experience professionals. The focus is on strategic technologies. We publish recommended articles and content technology news weekly. We do not sell or share personal data.

Subscribe | Feed | View online | Editorial policy | Privacy policy | Contact

© 2024 The Gilbane Advisor

Theme by Anders NorenUp ↑