Typically data that arrives in contact forms comes classified as part of the form completion process. Ventrica AI is designed to study and classify unstructured contacts.
We build AI classification models based on contacts previously classified. So, for example for one of the world’s leading fast-food restaurants, we trained a classification model on 50K email tickets.
When a new email/text/ chat/social comes in the data pipeline runs the message through the classification model and updates the ticket with the classification. The accuracy is greater than that of an advisor, even on contentious/split/multi-intent classifications. NLU Natural Language Understanding classifies the data on arrival and tracks sentiment (happy/angry).
Ventrica AI’s classification can be leveraged as a “first pass” triage to allow for prioritisation and skills routing.
It can also be used as a precursor to further automation (e.g. email with intent X always needs a product photo so if there is no photo attached ask for one automatically before the ticket reaches an advisor).
For voice contacts, we ask for the reason for calling in an IVR. We can then present the transcribed utterance to the advisor to add context to the conversation and then train on those utterances in the same way as we would for text-based channels.
Ventrica AI can apply sentiment analysis in the same manner.
Summarisation uses large language models (LLMs) to take a verbose ticket and summarise it into a few salient bullet points. It could just be a summary of the initial contact, or it can be an issue/resolution summary based on the entire conversation.
Ventrica AI uses summarisation so that business owners can see a more concise summary of each ticket rather than having to read through all the detail. It also gives advisors and managers a feel for the ticket before opening it. Agent assist benefits cannot be understated. Reducing handling time significantly.
Ventrica AI is a vast improvement on most software in the market because existing summarisation features only work after you’ve opened the ticket – and even then, you have to wait for it to summarise it real-time.
Ventrica AI is an extremely powerful tool for quickly turning disorganised knowledge data into something useful.
Combining knowledge management, semantic search, and generative Ventrica AI gives us an FAQ bot based on existing knowledge articles.
This is created using a Large language model that will interpret and store all customer data with guard-rails in place to protect both your data and your brand.
In this white paper, we look into why and how evolving customer preferences and advances in technology are transforming the customer experience.While we can’t all be Edison, Ford, or Bezos, we can certainly apply that type of thinking in our own roles to design new solutions for our customers and continue to improve existing ones.To do that, we need to understand what customers want today and how their preferences and demands are expected to change.
What major factors do we think will influence customers' preferences and behaviours in the future?Articles
See how omnichannel shopping can help bring together digital and in-person customer experiences.Articles
To be customer-centric, you need to consider the omnichannel service you provide.Articles