Tag Archive for: smart assistant

Posts

Portfolio Items

Pages

Customer support chatbot

265 billion customer support requests are made every year, and it costs businesses $1.3 trillion to service them.
Chatbots can reduce these costs significantly.

Request your free consultation

A customer support chatbot can engage users, reduce costs, provide 24/7 service, extend your reach and provide a business edge.

This is our example of a technical customer support chatbot for a fictional company that sells printers and scanners.

The customer support chatbot project


The customer support chatbot would be capable of answering the more common technical support queries related to the scanners and printers which the company sells.

The bot will provide an idea of what a fully functioning chatbot is capable of.

Highlighting the scope for integrating with existing back-end systems. In this case using a tech support ticket system, product inventory, and other internal system integration to answer customer queries.

This is a common use case for a chatbot and the solution could be deployed to a customer-facing website or onto a messaging platform. For example Facebook Messenger.

Support chatbot specifications


  • Lookup support ticket status
  • Provide product help
  • Create a ticket by integrating with internal ticketing systems
  • Arrange a callback with a customer
  • Answer common product questions
  • Provide in bot product information and links to further information
  • Connect directly with company knowledge bases to extract relevant information.
  • Hand over to live chat if necessary

An AI powered customer support chatbot


A customer support chatbot that can answer common questions and enquiries. Providing answers to common FAQ questions, the ability to look up and report on support ticket status, and provide printer support to troubleshoot common problems.

Leveraging machine learning and powered by advanced natural language processing technology.

Tailored to understand specific product lines and names.

Trained on a large data set of historical enquiries and capable of understanding 80% of user enquiries.

Using a mixture of natural language processing and rich UI elements to manage the conversational flow.

Ready to hand over if there is a drop in sentiment or a request for human interaction

Conversational Interface

Conversation Planning



The Chatbots aim was to support 3 technical support conversations.
Check Ticket Status
Get Help
Answer FAQ


Connecting to existing systems

Integration



The support chatbot has been developed to mimic a live application which has some back-end integration with existing systems.

The chatbot provides 3 levels of integration.

Customer support ticket lookup – the bot interacts with a current support ticket

Product listings – the user is able to drill down through product category choices when trying to get help for a specific problem.
Product lookup by serial number – the user provides a serial number which the chatbot uses to lookup the product details.

Platform

We decided on the Facebook Messenger platform for where our demo chatbot is going to reside.



Conversational Planning

3 Main conversational contexts.
Get help, Check current ticket, Ask FAQ. The bot also supports small talk by default, it will respond with predefined phrases. These can be tailored.



Understanding Users

Understanding the users’ needs, behavior, and expectations. This bot would serve current customers of the company requiring assistance with their printers/scanners.

Conversational Interface

We are able to leverage some of the more advanced UI elements in Facebook Messenger. Cards, Quick Replies, Images, Buttons.



Training

We can use the extensive training capabilities of the Dialogflow platform to fine tune the chatbot user experience.



Analytics

We can provide analytical data on chatbot sessions, queries per session, top intents handled by the agent and exit rate per intent.