Skip to content
Play Video

Natural voice and chat based

40-70% Agent Reduction

Multi-threaded Intelligence

Our vision

Customer Interaction Automation that works and conversational analytics that dynamically feeds knowledge and insight into every channel.

We believe current automation technology, whether chat and voice based does not automate or replace the agent, but merely manages a very small and defined subset of work. In many cases it just provides pre-conditioning before a human takes over. We estimate conversational automation is currently resolving less than 1% of customer interactions. Hence we still employ millions of human agents worldwide with the associated cost. We aim to solve this problem with CONNIE – The next generation of AI and NLP fuelled Agent replacement technology that will save companies Billions.

The problem

There are three. Today’s technology is both single threaded, but also based on pre-ordained decision trees. This means the automated assistants are only looking for defined questions or topics that have defined answers. Additionally it can only handle the right answers at the right time. If you are in a credit card chat or call queue and you ask, “do you have any mortgage offers right now?” the response will be very poor. Finally, the clients we serve demand highly secure, private and trusted platforms that operate at scale. Many competitors in the field rely on third party software to fulfil certain parts of the workflow. These are typically served from vendors such as Google, Microsoft or AWS. In every case, our current clients would not entertain using these platforms for this purpose and want a solution that is within their own environment. We have the entire IP stack, and can place this anywhere.

The Solution

We are building a next generation automated assistant that can respond naturally via voice or chat, in a multi-threaded conversation. In parallel, we leverage our domain leadership of interaction analytics in both human-to-human and human-to-bot use cases to constantly train and update our AI models for optimal accuracy and FCR – 24 x 7.