How AI and ML can translate data into wisdom By Aron Heller | Nokia, Monday, 07 Dec 2020

Will Smith’s I, Robot sees machine algorithms evolving to control humanity and inherit the earth. Ex Machina’s humanoid robot, Eva, betrays and kills her creator and hapless savior, Caleb. And Arnold Schwarzenegger’s Terminator is a cyborg assassin sent back in time to kill Sarah Connor, preventing her from giving birth to a son (who will eventually lead rebels against the machines that have taken over a post-apocalyptic future).

In the world of science fiction, artificial intelligence and machine learning are often the agents of evil. In reality, AI and ML are mainstream technologies that aim not to destroy mankind but to help us navigate the massive amount of information that dominates our lives. To paraphrase another hit movie: Data is “like a box of chocolates. You never know what you’re gonna get.”

How AI/ML helps make sense of data

Just as having lots of books doesn’t make you smart, having lots of data doesn’t lead to knowledge. The ability to extract and curate that data is what leads to real insight and discovery.

That’s where AI-powered algorithms and machine learning technologies come in. Companies like Google and Facebook have been at the forefront of using sophisticated algorithms to reach incredibly high levels of personalization from this sea of data, effectively leveraging machine learning for huge profits. Compared to Big Tech, telecoms unfortunately still have a long way to go in this regard.

Telecoms are losing the data race

When it comes to translating data to wisdom, telecoms have fallen short. Service providers may have access to massive information about their customers, but unlike Google or Facebook, they’ve been unable to construct and predict a holistic view of the individual’s wants and needs. This can result in an inconsistent and disjointed customer experience, – not to mention an enormous wasted opportunity.

Why have telcos struggled to leverage their customer data? First, much of the data resides in various siloed systems that offer solutions from different perspectives. Second, many of these are old legacy systems, with no way to connect the dots between them.

Complicating matters, communications service providers also face strict data protection regulations that limit their ability to use AI/ML on personal data.

Finally, there are many vendors that offer general AI/ML algorithms that are not telco-specific and could have potential side effects. Using such algorithms from many data sources without truly understanding the telco data could create an interpretation of an image of a dog being labeled as a cat.

Customers today expect service providers to give them a personalized digital experience, like Uber or Netflix. But we can only do that with accurate data in real time to give a full view, rather than just part of the picture.

How Communication Service Providers can leverage AI/ML

To begin with, we need to make sure that we can tap into all the systems, networks and OSS/BSS, new and legacy models alike, to get a complete view. The data is the ore and we need to mine it.
To extract information thoughtfully, we need to concentrate all the data in one place as a single source of truth. AI/ML algorithms can fuse the data, making sense of it and ciphering information that is telco-specific.
Additionally, AI and ML tools can help us deliver proactive care, network automation or marketing campaigns.

The business case for a holistic view

Data is the fuel on which businesses run. When the sheer scope of data is so staggering, advanced tools are needed to sort through and make sense of all those data points.

The only way to get a clear, holistic picture is by providing a complete view across multiple domains, bringing together an AI-driven view of customers, services, networks, billings and transactions to provide a common language across business functions. Only then will telco companies finally see themselves as their customers do: as a single company dedicated to a single goal.

That’s the potential of AI/ML. It’s not a tool to end mankind, but a way to help us become more efficient, more productive and more in tune with people’s needs.