The customer is a leading US insurance and financial services provider with major offerings in the life and auto insurance space. They serve over 30 million policies in the US and North America.

Business Challenges

Current state scenario involved the usage of emails and IVR based phone network for customer service, resulting in delayed response time and unavailability of customised responses/quotes. The customer wanted a scalable and inexpensive bot that would provide a highly personalised experience instantly, while having the ability to manage user information and data securely. The bot should also simplify claims and query resolution process for end users. The main goal was to enable customer acquisition across various channels especially where millennials are spending time.

We wanted to leverage bots because we want to reach millennials on various platforms. Velotio’s engineers were technically spot on and also recommended best practices and features that ultimately helped us a lot.
— VP, Digital Transformation

How Velotio Helped?

Velotio setup a 5-member engineering team that developed multi-channel bots which were deployed on Alexa, Facebook and client’s website. Our team tailored the attributes and tone of the bot to cater to the target market of the client. The bot leveraged Natural Language Processing backed by Artificial Intelligence to craft personalised insurance quotes for customers right from within the bot window and was launched within 11 weeks.

The bot was designed to offer 24/7 customer assistance and features included:

  • Getting customised quote for various types of insurance
  • Check the status of your insurance and renew it
  • Registering a claim
  • Checking claim status
  • Checking policy status
  • Address coverage queries
  • FAQs

Key Technologies & Platforms


The Solution

Two main components of the solution included:

1. Rasa NLU: Helped the bot interact with humans in the way that the humans find it natural. The bot was trained to handle over 110 intents (indications that tells the bot what the user would like to do) using RASA NLU along with hundreds of entities (Attributes which gives the bot details about what the task is related to). Example of an intent and entity shown below:

2. Botkit: Provided the provisioning to develop the bot once and deploy it on the two initially chosen platforms (Facebook and client’s website). We also integrated the bot with backend systems like the client’s CRM, database and core systems via APIs in order to get the quotes and details.

We can help you to build intelligent voice and text conversational bots within 30 days.
Get in touch to start a conversation.

Get in touch