It’s All In The Data
Customer Service is a treasure trove of data. Yet when confronted with such a wealth of information, most AI systems fail. There’s a way to unlock its secrets though: an AI technology called Clustering.
Why So Many AI Customer Service Projects Fail
We speak with many Heads of Customer Service who are all looking to AI to help them deal with: growing inquiry volumes, ever-higher standards of Customer Experience, and the struggle to cope with high agent-churn rate.
There’s a problem though: many AIs don’t perform well in Customer Service. The reason for this is straightforward: the AI has to learn so many questions that are being asked, in so many different ways.
It’s perfectly normal for a Customer Service centre to have relatively little overview on the breakdown of inquiries they might receive. They’ll have the high-level picture — bills, invoices, parcel tracking etc — but they won’t have the granular level of data necessary to provide answers.
Traditional implementation teams consequently have to manually guess what questions will be asked, and to differentiate between the many ways the questions are formulated. Such a deployment is extremely labour intensive.
Yet this data is vital to a successful and rapid AI implementation. Take the chatbot scenario. Natural Language Processing (NLP) is not a particularly new technology, and many NLP systems achieve an accuracy level of above 95%. Yet there are many situations where a simple chatbot will not understand you. This is not because the tech itself is in any way ‘bad’ — but it simply does not have the data range or experience with exceptions to answer your question.
Not everything can be guessed.
Data Becomes A Treasure Trove
This guesswork is the problem with simulation of data. A human might be able to guess at 100 different ways of phrasing a question. But in order to deliver an exceptional Customer Experience, a system requires multiple thousands of expressions in order to reply accurately.
To give an example from one of ultimate.ai’s clients, a Nordic airline:
One question might have 8000 ways of being asked.
There is a way that is is possible to avoid this manual data simulation exercise. Every customer interaction is already logged and transcribed completely. We just need a way to unlock that data, structure it, and use it to train AI to replicate the work of a customer service agent.
To do this, we use clustering.
Clustering is a deep learning technique. We use it to automatically classify millions of lines of unstructured conversations. In one day, it can bring an enterprise’s customer service data to life, revealing common cases, best responses, frequency, urgency and processes.
A service centre’s unstructured data is run through our proprietary deep learning algorithms, automatically clustered by the AI, and then implemented in an engine. Certain recurring cases (e.g. parcel tracking) can be entirely automated. The AI continuously learns from the data and the more it learns, the better it performs.
We use clustering to find the correlation between what customers say, and what they want. We can then provide the right answers according to need. Clustered groups can consist of huge data ranges and consequently create a far larger variety of expressions.
Through clustering, the ultimate.ai platform can go live in a month. The algorithms automatically create training data for the AI, meaning no manual work is needed. Not only is the deployment far faster with a clustered approach — the AI is also smarter. It understands customer need at the granular level and improves the overall Customer Experience. Language presents no barrier.
Clustering completely removes the burden from the client. The technology tells us what customers are asking, and offers accurate answers. The AI will constantly source expressions, and it will always outperform a human simulation.
Customer service data becomes a treasure trove.
The Future of Customer Service
ultimate.ai dedicates a lot of resources to clustering. It allows us to be language agnostic and scale across territories, it gives us very high suggestion accuracy, and it helps us to slash critical service metrics like Average Handle Time and Time to First Response.
Seeing the impact that deep learning, clustered AI can have on Customer Service is astounding. We believe that this is the future: technology that can deal with the volume, and that not only maintains Customer Experience, but actively improves it.
It’s taken us a long time to build, and much like our AI, we’ll never stop learning. We are helping some of the world’s leading brands deliver exceptional customer experiences every day.
Stay in the loop
Subscribe to our email newsletter. No spam, just occasional insight from our experts.