Skip to main content
Using Machine Learning to Improve Telecom Business Assurance

How can machine learning and advanced analytics improve business assurance in telecommunications?

That was part of the challenge we sought to solve in our first Catalyst project with TM Forum, which was showcased at Digital Transformation World Nice. Catalysts are proof-of-concept projects developed collaboratively by TM Forum members. These projects bring together companies large and small to create innovative solutions to common challenges, demonstrating how this can be achieved leveraging key TM Forum best practices and standards.

Working with Orange, Deutsche Telekom, British Telecom and Croatia Telekom, plus Amdocs, IBM and Bulb Technologies, we embarked on an adventure to address a common challenge within telecom business assurance. According to industry reports, revenue loss due to leakage, churn and fraud accounts for around 4% of total DSP (Digital Service Providers) revenue despite existing business assurance systems already in place.

Fighting Fraud and Controlling Risk in New Service Requests

At the point where a new customer orders new devices and / or voice, data, OTT services, DSPs need to make a risk assessment for credit, bad debt and fraud in order to understand the risk to the balance and balance against profitability. The telecom business assurance community needs to adopt AI, machine learning and open APIs in order to support DSPs in terms of revenue and customer experience.

The solution developed by our Catalyst team enhanced traditional credit and fraud assessments through the introduction of machine learning technologies in order to provide a more accurate and personalised risk assessment. This included:

  • Providing alternative and credit appropriate offers to those customers that were unable to meet the required threshold in order to retain customers.
  • Network link analysis in order to proactively identify suspect fraud.
  • Creating an open API-driven business assurance marketplace to help service providers pick the right partners and adopt new technologies without the need to upgrade all of their legacy solutions.
  • Contributing new APIs to TM Forum and new uses cases to revenue assurance and fraud management.

The project team demonstrated that if the APIs developed are applied in a real-world scenario, they could contribute to at least 1% improvement in the overall bottom line revenue for CSPs. This is a significant impact in an industry that is constantly fighting hard to control cost whilst investing in the latest technology. We even won an award for Outstanding Business Impact Catalyst — congratulations and thank you to all our partners on the project!

related posts