Smart Chatbot Case Study

One top-notch prospecting and client management solution used to maximize the effectiveness of customer engagement.

Overview

Pluaris Resolve brings together the core functionalities of Nowigence’s signature AI Engine, Pluaris, known for its industry-leading machine learning (ML) and natural language processing (NLP) techniques. Pluaris Resolve has been specifically designed as an advanced chatbot built to understand customer questions from a linguistic level, in other words Pluaris Resolve is more human-like than typical chatbots available on the market today.

The Problem

To date, it’s frustrating to deal with Chatbots. Either they are pre-scripted with answers to a few questions or operate with great limitations to address your concerns. Often times, we’ve heard that customers are repeatedly thumbing “0” or “9” to get to a human agent. After trying repeatedly and/or a long wait, the customers express their frustration in speaking with anger or providing cold answers. At which time “AI” Chatbots claim that they do sentiment analysis, which is foolish since the very reason that the customer is frustrated is because of the lack or inadequate response by the Chatbot.

The customer ask was simple, “Show me a Smart Chatbot that can answer any question if integrated with my knowledge base.”

smart chatbot conversation
AI for education administration

The Ask

Mr. Hassan A. El Shawarby, CEO of Athear (athear.com) reached out to Nowigence looking for a human-like “brain” as a virtual agent to help streamline his inbound and outbound customer service operation, currently serving global Fortune 1000 companies. He asked his team to send Athear’s own documents – SOPs, customer service manuals a day before the demo and wrote, “We look forward to having a successful call. We’re all geared up with some cunning and complex questions simulating real-time customer transactions.”

The Process

The following data processing steps were performed during the call with the Athear team, and they were shown the back-end processing in Pluaris as well as the outputs within Resolve. See Appendix for screenshots.

  1. Upload Customer Data: We were given roughly 10 documents to upload ranging from 500 words to 10 pages, including tables embedded within documents.
  2. High Accuracy Answers: In a flash their team was able to answer customer-like questions within Resolve. Answers were retrieved within the front-end and Athear’s team could see the same question in the back-end with varying confidence scores as well as additional supporting information related to the search topic.
  3. Seamless Knowledge Sharing: We then simulated when information is changed or updated and how quickly this permeates to the chatbot and the knowledge bank. When a different customer asked a question about the new information it was immediately available!

 

Integrations

During the call data was uploaded manually, and we didn’t connect to any of their internal systems. However, Pluaris Resolve can integrate with data sources like CRMs, billing systems, customer issue resolution systems, shared drives, knowledge banks, and other data applications that are required to address customer issues. These are all connected through a REST API integration.

customer experience platform sequencing

Before Pluaris Resolve

Contemporary products can answer limited questions or are preloaded with pre-scripted answers e.g., FAQs.

Questions are often asked repeatedly and if the pre-defined sequence of keywords are not present in the question the customer never receives an answer and can become frustrated or even buy from someone else.

Current chatbots do not answers questions off tabulated or structured data.

When current solutions connect with large databases they typically pull in many versions of the same document, which means you could be giving customers old information.

Most chatbot solutions have recognized that keyword search is being replaced by NLP techniques for more advanced searching and promise to deliver this capability within two years.

After Pluaris Resolve

Provides precise answers within seconds to specific questions with supporting “linked” information.

Resolve is powered by a linguistic engine capable of interpreting human languages and it breaks down every question syntactically instead of by keywords; which allows for an advance degree of freedom when customers ask questions.

Resolve can find answers within tables and structured or unstructured data.

Resolve only keeps the most updated intelligence so that your customers and agents are always getting the most relevant, critical information.

Resolve is already equipped with industry leading NLP techniques as well as linguistic comprehension, so your customers receive a more human-like experience when getting answers to their questions.

Conclusion

The Results

Pluaris Resolve addresses over 75% of customer issues and does it in a fraction of the time compared to contemporary products. What we’ve found is that when we equip agents and bots with access to more knowledge, their resolution times go down immediately. More importantly, when this intelligence can be accessed within seconds they deliver even better results.

AI for BFSI organizations

60%

Time saved by customers

50%

Increase to agent productivity

25%

Lower operating cost in first year

A Powerful AI Platform

what our customers say

Our ready-made solutions sync with your workflow.

No coding or specialized skills required!

trusted solutions. tested results. innovation that scales.
Contact Nowigence about AI software