Header
story

On the pulse of realtime

“Time doesn't go faster than it used to, but we run past it faster” George Orwell.

Each of us has this experience regularly. We want or have to communicate with a service provider and inevitably end up in an impersonal and almost always time-consuming and nerve-wracking Question Time. But why do we have to explain to the same company again and again who we are, what the status quo is and what we have in mind? Shouldn't it be a matter of course that we are recognized when we ring the bell or when we start our short message? Naturally. Since this is almost an art in itself, data scientists like Tomasz Sekulski work every day at Deutsche Telekom's Innovation Hub to personalize every interaction with our customers. But to do this, "machine learning" must work in real time and in parallel with the interaction with the customer and the whole process must take place at an average speed of less than 200ms – as fast as a blink of an eye. But that's not all, because the technology should really understand people and ideally also react to their emotions. That is why we at One Conversation work continuously with our team on the "Best Next Action Orchestrator" – a conductor, so to speak, who always keeps an eye on the entire orchestra and tailors each use of the instruments to the personal needs of the listener.

With this work, machine learning and “deployment phases” are interwoven a little more closely each day and thus enable a maximum positive experience when interacting – no matter when, no matter what topic and no matter on which channel.