4 recommendations for making full use of Big Data and AI

On 08-01-2021
Reading time : 4 minutes

Data analytics can give telcos and other service providers a competitive advantage and meet the needs of an increasingly demanding consumer. It offers insights into what is working well, where improvements can be made, and potentially, new segments to exploit. And AI can take it to another level.

Service providers have long looked for ways to get more out of data. Today’s new subscribers are tech-savvy consumers who have high expectations of the personalized digital interactions they want from their mobile, video, and broadband providers. The pressure is on service providers to quickly identify and meet these customers’ demands to remain competitive.

In addition to helping service providers develop a rich, real-time view of customers and their behavior, data can give insights into how to improve internal processes, cut costs, increase revenues, enhance efficiency, and more. But how can service providers extract actionable insights that create real value and then reduce the time between insights and actions? With all that data to manage, how do you store it, analyze it, and make the best use of it?

How can you get the most out of your data with AI?

Based on our extensive experience, Orange has identified four best practices that can help you unlock the potential of big data and AI.

1. Focus on shared data value chains

You need to focus data collection around shared data value chains and organize data by categories relevant to your use case. Ensure that you involve stakeholders that understand it and are familiar with working with that kind of data. Filter out data that is less relevant to your needs before centralizing it on a secure, scalable platform.

2. Choose the right infrastructure

25 ORANGE ILLUSTRATIONS AI PREDICTION W RVBHandling huge volumes of data requires the right infrastructure, but also means asking yourself some key questions: how much budget and time do you have available to allocate to formulating your infrastructure strategy? How often do you expect you will use the infrastructure? Do you have the necessary skills in-house to develop your infrastructure, or will you need to partner with a third-party expert? And if your data is sensitive, what sort of cybersecurity will you need?

The different varieties of cloud available can help you answer these questions. Cloud can give you the scalable, highly available, secured infrastructure you need to store large volumes of data, and has the flexibility to meet your budget requirements too. There are benefits to each type of cloud: public cloud lets you deploy advanced applications like AI without needing to own the infrastructure they run on and you only pay for what you use. If you want greater control and flexibility with data stored in your own data centers, a private cloud could be the right choice. For a blend of flexibility plus simplicity, a hybrid cloud may be your best option.

3. Deploy the right teams of people

Maximizing your data using AI and machine learning (ML) requires a mix of the right skills and people. Traditionally, operational teams are those people within an organization who field data and understand it at the point of receipt – but these workers aren’t qualified in AI knowledge and data intelligence. You should seek to form mini-ecosystems of people within your organization to exploit data to its maximum. Operational teams working and collaborating alongside data scientists who bring the AI expertise to the table and IT specialists to provide the infrastructure know-how can give you the right blend of skills and knowledge.

4. Plan for tomorrow today

Many service providers have used Big Data successfully on a few projects, but there is still the challenge of scaling it up. It’s, therefore, advisable to be proactive about testing use cases. Investigate what tools can let you implement viable use cases at scale: new AI platforms are emerging all the time that offer data science functionality and capabilities that will enable your teams, even without any deep knowledge of how to code, to conceive, test and deploy use cases quickly and easily.

A world of opportunity

The mutually beneficial nature of the data/AI relationship can give service providers an edge. According to Harvard Business Review, AI will add more than $13 trillion to the global economy over the next decade. With the right tools and skills in place, and by adhering to best practices, telcos and service providers can reap the rewards.

Orange knows data and AI

Orange has extensive experience in data analytics and AI, including 30 years’ research in neural networks, over 500 patents in AI and Big Data, and 2,200 experts working in AI and Data. We leverage AI and data to make smarter, more adaptive networks; to enhance the customer experience; and to improve our operational efficiency.
Below are some examples of Orange initiatives that have utilized data and AI to help generate new additional revenues and/or operational savings:

Fraud detection on an international voice network

Orionis, Orange’s AI-based voice fraud detection solution, analyzes 400 Gigabytes of data every day in order to identify fraud attempts in real-time.

Performance management of mobile radio networks

A new Orange AI solution can predict network congestion 30 minutes before it occurs. This innovation helps engineers save valuable time and proactively solve problems before they occur.

Optimization of mobile subscription procedures

Orange Senegal has developed a new, tailored solution based on robotic process automation (RPA) and AI to check subscription files from mobile offers, reducing repetitive tasks in its back office.

Optimization of mobile network investments

Orange Spain combines commercial and technical KPI analysis within a machine learning tool to optimize the rollout of mobile network antennas.

Enabling customized offerings

Orange has developed enablers for the centralized, responsible, and regulatory-compliant cross-collection of profiling data from multiple sources to assist with the personalization of retail offers.


You may also be interested in these articles: