Chatbot in B2B sales for Curandus

Curandus, a Boston-based medical startup, decided to check a chatbot for collecting leads on the website. We have created a chatbot based on Amazon Lex, AWS machine learning technology. The chatbot talks to potential clients, doctors and patients.

Chatbot in B2B sales for Curandus - image 1

Challenge: choosing the right technology for chatbot

The main challenge in a chatbot design process is the right choice of technology. We chose Amazon Lex due to its wide possibilities of creating conversation scenarios, the ability to learn (the bot saves every conversation with the user, and we then teach it), beautiful US English, the possibility of implementing other languages ​​in the future and the freedom to distribute the chatbot using known applications - Slack, Facebook Messenger or a website or mobile application. This means that our chatbot can talk to you even on Slack. Using Amazon Lex technology requires knowledge of Python and the ability to use the AWS Lambda cloud service.

Challenge: choosing the right technology for chatbot - image

Conversation with the chatbot: roles and scenarios

At the beginning of the conversation, our chatbot determines whether it is talking to a client (pharmaceutical company), doctor or patient. After determining who the interlocutor is, the chatbot launches appropriate conversation scenarios. The entire solution architecture is based on Amazon AWS technologies. The user is talking to the chatbot on the Curandus website.

Chatbot in B2B sales for Curandus - image 2

Conversation with Curandus chatbot

Talk to our chatbot. Remember to say hi and introduce who you are.

Brand colors

Primary variant 1
Primary variant 2


The chatbot interface allows you to use any font. We chose Nunito from the Google Fonts library, which is the signature font of the Curandus brand.

Typography - image
Chatbot in B2B sales for Curandus - image 3

Project stack

Date of publication:
27 Nov 2019
Done in 10 days
Find out more about the technologies we work with Technologies


Technical description


We did not have ready-made chatbot scenarios. The client provided us only with the structure of the data, which the chatbot was to collect from different types of users and deliver in the form of leads. Based on this data, we developed efficient and simple scenarios tailored to the requirements of Amazon Lex technology.

Chatbot design

For an experienced development team, implementing a chatbot from the programming side based on Amazon Lex is relatively simple. But designing the chatbot conversation flow correctly can be challenging. A chatbot is not simply a form that specifies exactly what data a user must fill in. The collection of data by a chatbot is done through a conversation, the design of which lies with the UX Designer. In a conversation, we cannot predict exactly how it will go and what exactly the user will give us. Please visit the blog and read more about chatbot design.

AWS tools

Lex, S3, DynamoDB, Lambda, IAM, Cognito, Amplify


Django, React.js, SASS, AWS