Chatbot - ValueLabs

End-to-End Case Study

Geo location

Europe

Platform

Web

Engagement

8 Years

Service

AI&ML

Challenges

The client faced a sudden surge in inquiries when COVID-19 hit early 2020. They wanted a self-learning chatbot for their corporate website to replace the live chat option.

Inconsistent

Unsupportive

Inefficient

Counterproductive

Deploy a chatbot on Atlas
Provide tool support and advice support to end users
Add skills to the bot to transform it into a virtual assistant
Provide support to restricted users through multiple channels (emails and phone)

Our Solution

For chatbot implementation, we integrated the chatbot with Microsoft QnA Maker allowing training and building a Q&A knowledge bank. We merged it with ValueLabs’ in-house built chatbot DAN to add Natural Language Processing (NLP) based search on unstructured documents. We developed a framework to extract answers from either DAN or QnA Makers, depending on the confidence score.

Solution Impact

Successfully addressed queries

Raised through chat, previously handled through support tickets

Improved response consistency

And customer satisfaction

Eliminated user restrictions

And enabled support to all the end users

Immersive Learning Experience: Enabling Growth through better Learner Engagement

Read Blog

Transform chatbot into self-learning virtual assistant

Previously, our client used to handle all queries from the end clients manually. However, they wanted to move towards chatbots. As an AI-powered chatbot provider, we were building a prototype when the COVID-induced pandemic hit, and the need for an intelligent and self-learning virtual assistant was at an all-time high. Our technical team brainstormed to identify the tools and technologies required to convert the proof of concept into a fully-fledged working bot in the shortest duration possible. Post-implementation, we planned for the bot’s knowledge base to grow continually. The goal was to reduce human dependency as much as possible while providing a consistently good experience to the end customer.

Deliver consistent customer experience despite bandwidth crunch

The legal service provider’s priority has always been fulfilling the end client’s needs. They have adapted their business operations accordingly and have been able to deliver exceptional customer service. Post the COVID-19 pandemic, the status quo changed, and end clients needed more help than ever.

Our client’s contact center received six times higher calls than standard rates, and questions ranged from furloughing employees to making the workplace COVID-safe. Our client struggled to answer the queries on time and risked losing business. ValueLabs stepped in with an innovative solution: an intelligent, self-learning chatbot for the client portal. The bot was live within two weeks and was immediately able to resolve one-half of the questions asked. The figure is now at two-thirds as the bot continues to improve itself.

Deliver consistent customer experience despite bandwidth crunch

Reduce human interference and maximize usage of NLP

The chatbot reacts like a human agent would by using AI and NLP to interpret the questions posed. It does not require any dedicated configuration or training. The AI-NLP chatbot can promptly find solutions in various resources, including text documents, movies, and graphics. We developed it to pick up new information from every interaction. Amid the COVID-19 crisis in 2020, the client was able to offer quick, knowledgeable, and friendly customer service. The impact of our efforts was visible as a significant growth in the client’s Net Promoter Score (NPS), as it grew by ten points between February and December 2020, during one of the most challenging times for businesses worldwide.

Business Impact

When COVID-19 hit, the UK-based professional services firm experienced a 6x increase in inquiries and struggled to handle the volumes of inbound calls. We developed an intelligent, AI-powered chatbot with powerful neural search capabilities and NLP for a conversational experience.