INTRODUCTION
The VA solution can be deployed through both voice and messaging channels, helping to provide enhanced self-service interactions that engage every caller in natural language, not the stereotypical, robotic language that is often associated with new technologies.
On this theme of improving customer experience, the Virtual Assistant will also enable you to offer 24-hour support, ensuring business continuity – especially amongst organizations with high call volumes and/ or text interactions.
While the contact center queue is peaking, simple tasks can be offset by the AI-driven conversations and voice-enabled self-service capabilities of Mediatel Data Virtual Assistant, as first call resolution increased by 20% in many cases.
Not only this, but business insights will be increased – while automating either the entire or at least the beginning of the conversations, things become a little easier for the agents. Calls to agents that require user authentication have reduced their time by 40 seconds per call, due to authentication by Virtual Assistant.
All of this drives down costs, as agents have to handle fewer calls and – when they are called into action – will have lower handling times. It’s important to mentions that the VA and agents have a 360* view on all interactions a customer had, offering them a precious feeling of unified communications, when the integration is made with Mediatel Data Omni-Channel Platform.
For the customer, wait times are reduced, straightforward queries are immediately resolved and customer satisfaction will likely be increased. Our latest internal market research showed that customer intentions are identified with an accuracy of over 95% with Mediatel Data VA.
The solution is designed to improve business, customer and agents results, providing a significant decrease in the need for agents to return customer in IVR and increasing overall customer satisfaction.
Virtual Assistant through Mediatel Data Business Model
During the past year, Mediatel Data has laid the foundation of an Artificial Intelligence department that develops smart solutions dedicated to the field in which we operate. The need for this department arose from the needs expressed by our customers, that have guided, as always so far, our decisions and products.
The role of the AI Department is to boost practical applications with research-inspired ideas. By working closely with the development and support departments, we have designed and implemented Machine Learning-based modules for the Mediatel Data Call & Contact Center software, all gathered under the concept of Virtual Assistant.
Ranging from identifying the intents of customers that contact a business via several channels (voice call, video call, e-mail, WhatsApp, Facebook, etc.), understanding the topics that they are interested in, analyzing their attitudes or proving their identity (via voice biometrics or facial recognition), we have integrated numerous new technologies, as part of our continuous efforts to offer the best products to our clients.
The Mediatel Data Virtual Assistant has already been embraced by some of our customers. This ensured access to production data, thus allowing us to train from the very beginning accurate and effective AI models.
Moreover, due to the dynamicity imposed by 2020, we have challenged ourselves and went even one a step further, by developing 100% personalized solutions in a quick yet professional manner: in just one year, we have succeeded in customizing our AI solutions for banking, accounting, telephony, energy management, insurances and medical services.
Our AI modules are entirely developed in-house: trained on our systems—not on proprietary cloud platforms, with real data that we collect from our clients. Thus, the client is not obliged to buy third-party software, the models are more flexible and can be easily adapted to the clients’ needs. Finally, the deployment can be 100% on-premise, on the clients’ servers, thus complying even with the most severe security constraints.
Secondly, our Machine Learning models are dynamic. Most companies offer pre-trained models that can solve fixed tasks, without investing any effort in improving them over time. Our Virtual Assistant, however, is designed to allow constant improvements. When landing a model in production, we accompany it by feedback mechanisms through which the users can signal the misclassifications. This information is then used to periodically re-train the model. Hence, we make sure that the models are always complying with the most up-to-date business needs.
Last but not at least, it has become popular among businesses to offer a “good at everything” assistant that has the same core tied to different input/ output sources (text, video, voice).
Nevertheless, the Mediatel Data Virtual Assistant is never personalized for a customer by having a single workflow in mind: the performance of our product lines in our awareness of the fact that different channels require separate flows. For example, we have specialized models for both written and spoken text—and that’s how we succeed in capturing both basic and subtle features.
