Databricks Inc. today launched another in its series of vertically oriented analytics and artificial intelligence platforms, this one targeting telecommunications carriers and network service providers.

The Data Intelligence Platform for Communications combines the company’s data lakehouse with large language model features Databricks picked up with its acquisition of MosaicML Inc. last summer. MosaicML provides an LLM that companies can train and fine-tune with their own data at low cost.

Databricks said the offering provides a unified foundation for data and AI development that gives service providers a holistic view of their networks, operations and customer interactions without sacrificing data privacy or risking disclosure of confidential information. It combines data management, governance and data sharing with generative AI and machine learning tools.

The company has been steadily building out its portfolio of vertical lakehouses since the launch of a platform for retailers two years ago. It also has industry-specific platforms for financial services, healthcare, manufacturing, public sector, retail and media.

Databricks said the communications platform can help communications service providers better forecast market trends, predict demand patterns, make money from data as a product, and make analytical insights available to a wide range of employees. AT&T Corp. has used more than 100 machine learning models into production for such purposes as fraud prevention and operational efficiency.

CSPs get a single and comprehensive view of their customers’ structured and unstructured data that they can use to reduce churn and deliver more personalized experiences. CSPs can govern all data types, machine learning models, notebooks, dashboards and files on any cloud using the Databricks Unity Catalog. The platform includes Delta Sharing, an open protocol for securely sharing data in real-time to enable partner collaboration.

Customers can also access pre-built Databricks and third-party “accelerators” that include LLM-powered chatbots, network analytics, geospatial analytics for fraud identification and customer entity resolution, which identifies, matches and merges records that correspond to the same customer across different databases or data sources.

Photo: ArtisticOperations/Pixabay