How A Semantic Layer simplifies your Data Architecture- ATSCALE

Anurag Singh
3 min readJan 6, 2022

Problem Statement

Making Data accessible to everyone in organization is a challenge which almost any organization under the sun faces. E.g. Data Scientists(DS) do forecasting/predictions Business Analysts(BA) manage and drive revenues. So the need for a Unified Fabric that joins together BA and DS community; does that in a user friendly business friendly way following consistent set of Business metrics and allows both parties to use the tools they already know meaning no learning curve is involved. It should also provide IT single source of truth and ability to govern data across organization.

So what basically is a Semantic Layer?

A Semantic Layer is business representation of corporate data that helps end users access data autonomously using common business terms. A Semantic Layer maps complex data into familiar business terms such as product, customer, revenue to offer a unified consolidated view of data across organization.

ATSCALE is one such find which maps complex data into familiar business terms for Data Science and Business Intelligence programs built on Cloud.

ATSCALE Properties

1.Presents a consistent set of business metrics for BI and Data Science teams to consume from with tools of their choice.

2.Establishes an integration layer within enterprise data fabric to support analytics discoverability, governance and security.

3.Accelerates end to end query performance while optimizing cloud resources.

So where does Semantic Layer Sit in your Analytics Stack?

*ATSCALE Whitepaper

The ATSCALE semantic layer sits between your analytics consumption tools and your data platforms. By abstracting away the physical form and location of data, the ATSCALE semantic layer makes data stored in data lakes or data warehouses accessible with the same interface. Integration with enterprise data catalogs makes ATSCALE models discoverable and metadata shared seamlessly.

Lets get into the architecture and how ATSCALE achieves so much.

The answer lies in the way components are designed to function

*ATSCALE Whitepaper

1) Consumption Integration- Anyone using Excel, Power BI, Looker or Tableau can connect to ATSCALE and run queries immediately. Application and integration friendly via REST, JDBC, ODBC, MDX and DAX.

2) Semantic Modeling- ATSCALE’s semantic model unifies semantic definitions and metrics for data and makes it available in one location for BI, AI/ML and applications. It works on data anywhere whether it’s in a data lake or a data warehouse

3) Query Virtualization- ATSCALE’s data virtualization automates the sourcing, curation and modeling of data on premises or in the cloud. It blends live data from multiple data sources into virtual, logical views. Virtualization makes IT more agile with the ability to store data in the most suitable platforms while providing the flexibility to adopt new platforms in the future.

4) Performance Optimization- ATSCALE’s autonomous performance optimization technology identifies query patterns and creates and manages intelligent aggregates, just like the data engineering team would do. The AI-driven optimizer learns from user behavior and data relationships and takes care of data updates and changes.

5) Analytics Governance- Enterprise Directory Integration (AD/Octa/OAuth) Role-based, object level control for users/groups and Column masking in query tools.

Data Integration- ATSCALE speaks to data lakes and data warehouses with data platform optimized SQL so performance is as fast or faster than hand-written queries. Rather than processing data locally, the ATSCALE engine pushes queries down to the underlying data platform to eliminate data movement and scale performance along with the data platform, without the need of managing a separate analytics infrastructure.

ATSCALE Query Engine-The ATSCALE Query Engine acts as a query interface for business intelligence, AI/ML tools and applications. Tools can connect to ATSCALE via one of the following protocols:

For the tools that speak SQL, the ATSCALE engine appears as a Hive SQL warehouse. For the tools that speak MDX or DAX, ATSCALE appears as a SQL Server Analysis Services (SSAS) cube. For applications speaking REST or Python, ATSCALE appears as a web service.

More details in their demo page



Anurag Singh

A visionary Gen AI, Data Science, Machine Learning, MLOPS and Big Data Leader/ Architect