Streamlining analytics workflows for CSPs drives efficiency and reduces costs

Summary

Omdia’s 2020 ICT Enterprise Insights survey indicates that more than 80% of customer service providers (CSPs) are focused on reducing operational costs and complexity in 2020. Analytics and data management tools are key capabilities they plan to invest in to address these challenges. CSPs, however, need to adopt a more efficient approach to running their analytics and artificial intelligence (AI) projects to maximize return on investment from these tools.

Investing in a common analytics platform, including a portfolio of applications, with integrated data pipelines to support these applications can help CSPs streamline analytics projects. Several vendors are developing solutions that leverage these capabilities, including Guavus, which recently launched its multiproduct Guavus-IQ portfolio to enable CSPs achieve network- and customer-related use cases, accelerate analytics workflows, and facilitate decision-making capabilities to address business needs. The deployment of these solutions will, however, benefit from the industry collectively addressing the data access challenge.

CSPs need to streamline data analytics assets and workflows

Increased costs and long project times are challenges CSPs face when performing analytics projects. Projects are often conducted by different teams within the CSP business collecting, storing, and analyzing similar datasets. For example, network- and customer-related data is collected and analyzed by different teams to serve multiple business objectives including accelerating new product or service launches and making networks more secure and efficient. Efforts and investments are duplicated, leading to unnecessary cost.

Data scientists also spend a lot of time cleaning datasets and preparing them in a format that is ready for analysis. These processes take the greatest amount of time in the analytics cycle, slowing down other processes that are required to access the data insights needed to address business needs. This delay is a major contributor to the failure of CSP digital transformation projects.

CSPs can address these challenges by investing in a common analytics platform with prepackaged analytics applications and automating the data pipelines associated with these applications. A common analytics platform can analyze the data collected and stored from the business to provide insights that meet the analytical needs of organizations across the business. This platform should, however, provide self-service analytics capabilities for business and IT teams to create more applications as the need arises. With such a platform, the different business teams only need to access a single platform to meet their analytical needs.

Automating the data pipelines of these analytics tools addresses the challenges data scientists face when preparing data for analysis. Data pipelines provide information on the relevant datasets and enable automated workflows to cleanse, format, and enrich the data and perform advanced functions, such as leveraging machine learning, on the collected datasets before they are passed to the applications. The automated data pipelines can also enable CSPs to quickly extend analytical use cases.

The industry requires such solutions to accelerate the use of analytics and AI to improve operations and revenue performance. Vendors such as VMware (based on its recently acquired asset Uhana) and Guavus offer solutions to help data scientists focus on developing analytics workflows and not waste time on the data acquisition and preparation stages. Omdia expects more vendors to do the same.

Guavus-IQ, an example of a common analytics platform using data pipelines to speed up analytics

Guavus’ recently launched Guavus-IQ product portfolio aligns with this solution. It is a portfolio of products that run on the Guavus Reflex platform. Guavus-IQ correlates and analyzes external and/or internal CSP datasets to address multiple analytics use cases within a CSP business. Two analytics applications currently run within Guavus-IQ:

• Ops-IQ includes fault and service experience analytics modules to help CSP network operations teams drive efficient operations and improve customer experience.

• Service-IQ was developed to address the needs of marketing and customer care teams by providing real-time insights into subscriber, device, and network behavior.

Guavus-IQ provides a preintegrated data pipeline used to develop these applications to reduce the time spent identifying and preparing the datasets required to execute the analytics applications. While Guavus currently offers CSPs support services to extend these applications, CSPs with in-house data scientists can utilize these data pipelines to extend functions of prepackaged analytics products in line with their needs. They can speed up access to insights and reduce associated costs. Reliance Jio is an example of a CSP that is currently utilizing Guavus-IQ and its prepackaged applications to support the development of its big data analytics and AI use cases.

Data pipelines are relevant but will be hampered by the data access challenge

While analytics vendors such as Guavus make data pipelines available to accelerate analytical processes, data pipelines are most effective when required data sources are readily accessible. Access to data remains a challenge for CSPs. Therefore, to gain maximum benefit from data pipelines, CSPs have a role to play in ensuring that defined data sources within data pipelines can be accessed.

There are attempts to address these data access challenges through the efforts of CSPs and vendors via organizations such as Open RAN, O-RAN, Open Core, and TM Forum. These organizations are defining open interfaces to ease access to data within CSP network infrastructure and supporting systems. The industry is, however, yet to align with specific interfaces that CSPs and vendors should adopt. It might take a while to resolve the data access challenge; it is, however, critical to successfully implementing AI and analytics in the CSP environment.