Databricks on Wednesday launched the Data Intelligence Platform for Communications, an industry-specific version of the vendor's tools targeted at telecommunications providers.
The Data Intelligence Platform for Communications is the sixth industry-specific version of Databricks' platform.
The first such version of the vendor's platform was introduced in January 2022 and geared toward retail enterprises. Since then, Databricks has also developed versions of its platform for companies in finance, healthcare and life sciences, manufacturing, and media and entertainment.
Each is made up of Databricks' general-purpose tools that have been tailored to meet the needs of given industry verticals. In addition, they include access to partner-built data products, governance capabilities, connectors, relevant data sets and other features geared toward organizations in a specific industry.
Among other vendors, Snowflake -- which launched a version of its data cloud designed for telecom providers in February 2023 -- Microsoft and SAS also offer industry-focused versions of their platforms.
The main benefits of such suites include simplification and improved efficiency, including faster onboarding given that enterprises don't have to alter general-purpose tools to meet the specific needs, according to David Menninger, an analyst at ISG's Ventana Research.
"All users of software platforms operate in a particular industry or perhaps in multiple industries," he said. "Industry-specific portfolio offerings recognize this reality and make it easier for organizations to adapt platforms or products to their needs. They are a great head start allowing them to be productive sooner than if they had to start from scratch."
Kevin Petrie, an analyst at Eckerson Group, likewise noted the improved time-to-value that platforms tailored to meet the needs of enterprises in different industries can provide.
"Data types and use cases can vary by industry, so when a software vendor in this space reaches a certain size, it makes sense to start codifying industry-specific models, templates and procedures into its offering," he said. "This helps adopters derive value from the software more rapidly."
In its rollout, Databricks noted that the communications industry is in a period of notable change. Global telecommunications traffic is on the rise, which requires adding more network equipment, and consumers are demanding higher-quality services and customer experiences.
Telecommunications providers, meanwhile, tend to be more data-driven than organizations in some other industries, Petrie noted. Therefore, given the rising need for analytics in telecommunications and the propensity for such companies to use data, Databricks was wise to target the industry with its newest customized platform.
"Telecom companies tend to be early adopters of new analytics technologies because data can give them competitive advantage in a price-sensitive, commoditized market," Petrie said. "For example, [they] optimize infrastructure performance and reduce customer churn."
Based in San Francisco, Databricks was one of the pioneers of the data lakehouse architecture for data management and analytics.
Lakehouses are essentially a combination of data warehouses, which store structured data, and data lakes, which store unstructured data. By combining the capabilities of data warehouses and data lakes in a single environment, lakehouses enable organizations to combine structured and unstructured data to gain a more comprehensive understanding of their operations.
As a result of its specialty, Databricks named its capability set the Lakehouse Platform.
Over the past year, however, the vendor has made enabling generative AI development a focus.
Toward that end, Databricks acquired MosaicML in June 2023 for $1.3 billion to help customers develop and train their own large language models, introduced LLM and GPU optimization capabilities in October 2023 to help users improve generative AI outcomes, and introduced a new suite of capabilities in December 2023 designed to enable users to customize generative AI applications with their own data.
Due to its new emphasis on generative AI along with its continued focus on the lakehouse architecture, Databricks unveiled the Data Intelligence Platform in mid-December.
The Data Intelligence Platform for Communications is Databricks' first industry-specific suite of tools to reflect the new name of the vendor's capabilities. In addition, it is the first industry-specific portfolio from Databricks to include generative AI capabilities at its launch.
Among other tools, the Data Intelligence Platform for Communications features LLM-powered chatbots aimed at augmenting human support teams and improving customer support.
"Customer service is a good place to start with LLM chatbots," Petrie said. "Customers have grown accustomed to chatbots over the last decade. If GenAI can improve chatbot accuracy and speed of resolution, that's a win-win for companies and customers alike."
David MenningerAnalyst, ISG's Ventana Research
However, given that LLMs sometimes hallucinate -- deliver incorrect responses -- there is some danger in trusting a chatbot to interact with customers, he continued.
"The risk, of course, is that LLMs can upset customers and hurt revenue if they hallucinate," Petrie said. "To reduce this risk, Databricks is making it easier for telecom companies to … prompt [the LLMs] with accurate, domain-specific data."
Menninger, meanwhile, noted that the LLM-based chatbots will enable more workers within organizations to work with data analytics.
Studies have shown that only about one-quarter of employees within most organizations use data and analytics in their workflow, in part because of the need to code to use most analytics tools. LLM-based chatbots enable data queries and other analytics tasks without requiring code.
"Generative AI is making data and analytics much more accessible to large portions of the workforce who may not otherwise be able to find the information they are looking for," Menninger said. "The majority of the workforce in the majority of organizations doesn't have access to analytics, in part, because the tools are too difficult to use. LLM-based chatbots will help address this issue."
Beyond the LLM-powered chatbots, the Data Intelligence Platform for Communications includes the following:
- Telco Network Analytics, a feature designed to improve network reliability as 5G gains popularity as well as reduce customer churn with forecasting capabilities.
- Geospatial analytics capabilities that analyze spatial movements at scale to identify fraud from unusual card transactions.
- Unified governance of all data types, machine learning models, dashboards and files on any cloud as well as built-in compliance through the Databricks Unity Catalog.
- Collaboration with Delta Sharing so telecommunication enterprises can securely share data and data products with partners without requiring those partners to invest in the same technology.
- A comprehensive view of customers by bringing together data generated from transactions, support calls and other events in a single location.
While each should benefit telecommunications companies, Telco Network Analytics has the potential to be the most significant of the non-GenAI capabilities, according to Menninger.
"Telco Network Analytics is very specifically focused on the telecom industry whereas [other tools] apply equally to other industry segments," he said. "Creating models that analyze call detail records and network performance measures allows communications companies to improve the reliability of their networks and improve customer experience, which, in turn, helps reduce customer churn."
Petrie similarly pointed to Telco Network Analytics as a significant tool within the Data Intelligence Platform for Communications. In addition, he said the comprehensive customer views are also important.
"The accelerators for network analytics and 360-degree customer views involve a wide range of multi-structured data, including tables, log files and text," Petrie said. "This underscores Databricks' commitment to serving as the unified platform for all data and analytics use cases, including business intelligence as well as its heritage strength in data science."
With generative AI now part of one of Databricks' industry-specific platforms, Menninger said he'd like to see Databricks add custom LLM capabilities to its platforms for other industry verticals.
When Databricks launched its Lakehouse for Retail in early 2022 and subsequent industry-specific lakehouses up through the Lakehouse for Manufacturing in April 2023, it hadn't yet acquired MosaicML or developed other capabilities aimed at helping customers develop and train generative AI applications.
The Data Intelligence Platform for Communications, therefore, is the first of Databricks' industry-specific platforms to include generative AI capabilities simply because of the timing of its release.
"Databricks was one of the first data platform providers to focus on enabling custom LLMs in their platform," Menninger said. "I expect we'll continue to see more of these capabilities across other domains. It's all about business value. If you can't combine the technology with the business context, you'll leave your customer disappointed."
Eric Avidon is a senior news writer for TechTarget Editorial and a journalist with more than 25 years of experience. He covers analytics and data management.