Last time we explored how the broader market adoption of AI technologies was unfolding and explored some of the biggest opportunities which lay ahead for enterprise grade AI, now we take a deeper look into Artificial Solutions Virtual Assistant platform Teneo, and how businesses can benefit from implementing.
AI Business: Can you start by explaining more on Artificial Solutions proposition and how the platform works?
Dave: Where we feel the interest – and benefit – lies for enterprises is in powering that level of human-like experience.
We like to describe this as powering natural language solutions which allow people to communicate with a product or service or with a device in free-format natural language or unstructured dialogue and through that it’s able to come across as a very human-like experience and what we’re essentially doing is, and we use the tagline ‘make technology think’ to describe this.
There are whole new generations of people that expect to be able to pick up a device, an application, a piece of software and to have in understand them – without needing any training. Thus we need to create things that are so simple they don’t require a lot of human learning, or more so that we need to actually train the technology, we need to make the technology think and adapt to how the user wants to interact with it.
Now behind the scenes, creating this highly intuitive – typically natural language enabled solution – takes a mass of sophisticated technology. And that’s what we do. We’ve packaged up our 15 year experience of building natural language applications across over 30 different languages, and we’ve packaged it up into a development and analytics platform that we call Teneo….Latin to know…. so that non-computational linguists can build these artificially intelligent applications – quickly!
AI Business: Could you give an insight into how it actually works in enterprise setting both customer facing and for internal applications?
Dave: Our underlying platform can be applied to these natural language interaction solutions, and that is the way we try to encompass the layer that we are adding in to these intelligent virtual assistant solutions. This natural language interaction layer can be injected into a variety of customer facing applications and devices from mobile personal assistants to intelligent online applications to speech-enabled consumer devices and it’s increasingly being used in a similar way in enterprises for customer service, online sales and speech-enabled enterprise apps.
And increasingly, this customer facing capability is being used to span the entire engagement cycle… starting from when people are considering buying a product or service to when they actually complete the purchase (“sales assistants” or “shopping assistants”) to the actual use and on-going support for that product or service.
It could be a technology product, an insurance policy, financial instrument or a home automation/security service – whatever that product or service is, then consumers want to have that same level of natural language, free-format interaction capability and when they have a question or an issue, they want to get intelligent, automated and informed resolution quickly and effectively.
AI Business: We’ve seen that Shell has recently adopted the Teneo platform, how are they making best use of its capabilities?
Dave: With Shell it’s not just direct to customer but it is also benefiting their partner relationships. I mean very much in Shell’s case, it’s distributors and it’s how distributors can get access to the most current set of information and recommendations of using the latest and greatest Shell products, so in this case, the customer for them isn’t always the consumer, it could also be the distribution channel.
AI Business: In terms of tech development, what do you see as being the biggest hurdles to get to a point where we are getting towards full adoption?
Dave: Building intelligent natural language applications is not an easy thing. Yes, you can do anything if you have an army of developers, unlimited resources and very deep pockets but in the past, it’s only been the likes of Apple, Google and Microsoft that can do this.
So in the past, when enterprises want to start building natural language applications they had limited options; either try to develop something themselves or go to a vendor who would hand-crank a natural language artefact just for them. We have taken a different approach. We’ve taken our 15 year experience and combined it with the immense corpus of data built from millions of real conversations in multiple languages. We’ve taken the best approaches to natural language processing combining both statistical and rules based approaches to NLP to build a hybrid model. And we’ve built an open architecture that allows specialist plug-ins to be added so if a client needs a named entity recognizer for their own domain expertise, it can be built once and re-used many times. And we’ve packaged this up into a patented platform that allows non-specialist to build these applications – quickly.
We believe that this open approach to providing a platform that allows our partners, clients and ultimately the wider developer community to build a whole range of natural language applications is the way to overcoming the inherent hurdles and delivering wider adoption.
AI Business: What is next for Artificial Solutions?
Dave: We’re already seeing what our platform can do for large enterprises with artificially intelligent natural language solutions that span both internal and external facing applications.
The next step is the ever-growing emphasis in our R&D around the natural language data and analytics ability and how we can help these enterprises take the most advantage of the wealth of knowledge locked up in all of these unstructured dialogues that’s coming in from their customers and prospects.
For a really good example of how this is helping organizations, take a look at the financial sector. For many years, they’ve focused on cost savings and have driven their customers away from any sort of interaction with initiatives such as internet banking, ATMs and other non-branch activities. The problem is, there is no a massive customer disconnect – banks no longer talk to their customers. They don’t know what they want. They don’t have a relationship with them. By building intelligent applications that customers can talk to – and still get the self-service that they now crave, allows banks to re-engage with their customers, providing a better customer service and understanding what is in the mind of their customer.
We believe this is the future of artificial intelligence in the enterprise.