CRMXchange Interviews Omilia CEO Dimitris Vassos

Omilia’s Omnichannel Virtual Assistant platform provides a seamless human-like conversational experience for self-service across all channels. It enables end-to-end natural language conversations with advanced cross channel continuity and hand off. Based in Greece, the international company is growing 100% year over year with 170 employees. Omilia has over 20 production call center deployments of true Conversational Virtual Agents in 11 countries worldwide, representing 10 different languages, and serving more than 1 billion customer interactions. It has partnerships with inContact, RingCentral, and Amazon Connect and offers expert packages pre-built for specific industries. such as banking, insurance, and healthcare. Dimitris Vassos, Founder and Chief Architect of Omilia, offered additional background.

On your website, you state that your conviction is that ‘good technology is invisible to the user.’ Can you tell us how you ensure that your solution is easy to implement and easy to use?

There is a misconception in the market that AI technology is all about a clever neural algorithm. It’s not. Algorithms do nothing without data. Data– not just any data– but the right kind of data, is what makes the difference between great tech in Powerpoint and a successful deployment within the enterprise.

We believe Omilia is the first AI vendor to integrate the entire AI technology stack into an enterprise-grade platform, and package it along with high-performing, pre-trained industry specific models, necessary for enterprise clients and partners to deliver conversational customer care just using intuitive tools with user-friendly GUI interfaces, getting 90% accuracy, across the entire domain, out-of-the-box.

Omilia’s virtual agents are not just easy to implement in record time; they perform so well that 60% of callers feel compelled to say “Thank you” before hanging up with the system–just like talking to a human.

How does your platform enable enterprise organizations to accelerate the digital transformation of their customer care?

Omilia has created enterprise-grade, easy-to-use, out-of-the-box working resources for fast development of conversational virtual agents specializing in specific vertical and language combinations (e.g. US EN – Banking, US EN – Insurance). These resources, which we call Xpert Packages, allow enterprise organizations to bootstrap highly effective virtual agents and only need to up train them on the delta between the domain and their own business specifics.

But apart from that, when you are deploying in an enterprise, such as a large financial institution, there are other, more mundane factors, aside from AI accuracy: things like PCI-DSS compliance, GDPR and PII. These are all showstoppers if you prove that you support them out-of-the-box. That’s the value of having a Platform, versus having a Technology. There is now one single system that packs all the understanding and works across all channels! Easy to maintain, and at the same time ensuring clients never have to do things multiple times to accommodate different channels.

Why do you consider it critical for companies to be able to conduct conversational customer service interactions across all channels?

The ideal customer experience allows customers to interact conversationally on the channel of their choice, all while maintaining the context of those interactions. Just like talking to a friend. It should not matter if you are talking or texting. The only thing that matters is getting the right kind of response effortlessly and in a conversational manner.

Brands that offer this type of omnichannel customer experience are keeping their customers happy and earning their loyalty. Brands that don’t, pay the price of being absent, because only a friend in need is a friend indeed.

In what ways do your solutions differ from similar offerings available on the marketplace?

The competition has created solutions that target app developers and the wider B2C market with simple and quick to market products/services. Their downside is that their Natural Language solution is trained on a very limited sample of developer supplied example requests that are then used by the system to categorize similar subsequent utterances into intent categories. This approach is best suited for creating impressively fast Proof of Concept grade virtual assistants that fail to scale and satisfy the minimum requirements of enterprise and SMB clients.

By contrast, Omilia trains its Natural Language AI on datasets of millions of actual user utterances for each vertical (Banking, Telco, Insurance, Healthcare) and language combination. We have thus developed and already deployed our proven enterprise grade platform in more than 15 countries and more than 30 enterprise customers. Omilia’s proven enterprise-grade platform allows for real, end-to-end conversational service, that actually exceeds human accuracy; in both semantic accuracy and intent classification.

Read the entire 2019 CRMXchange Enterprise Connect Review here, and the original interview here.