Back to blog

AI Chatbot for Automation Building Process


Project description

We have built a solution that allows users to get information about the construction process, stored in the Geometrid database. Its flexibility allows it to be easily used with different projects without making any major changes to the bot.

Duration: 1 month

Development team
Developers, Tech Lead, Project Manager, QA Engineer
What we built
The solution includes

Support for different messengers, like Facebook Messenger, Slack, Viber or Skype, as well as a web-based widget.

The services and frameworks we used

DialogFlow is the platform, which provides analyzing and recognition of user’s query, as well as easy integration with many messengers.

Node-RED is the programming tool for wiring together hardware devices, APIs and online services.

AWS Elastic Beanstalk is the service for deploying and scaling web applications and services

Development process

We always demonstrate to our clients how their platform or bot are developing to assure them of high-quality services and the team’s product set in mind.

Firstly, we’ve gathered all the client’s requirements and brain-stormed regarding best practices of implementation. When the strategy of project running was identified, we started to develop chatbot using DialogFlow. After that, we gathered all possible required intents and set up them all. We created the server with Node-RED help and set it up to process user’s questions into database queries. The last step, when the logic was created and tested, we deployed it to Elastic Beanstalk server.

Bot architecture
Bot architecture
The result

The Chatbots.Studio team built a solution that allows for easy and quick access to information stored in the database about any object related to the building. The main benefit is, that using this bot does not require the user to know any database querying languages, or install any additional software, as this bot is available on the most popular messengers.

Back to blog

AI Chatbot for Local Coffee Chain


Project description

The main solution idea is to create closer relationships between coffee lovers and cafe chain representatives. Since the chatbot was launched, clients have simple access to special offers and discounts. Before, this customer can log in with a phone number. When the customer is looking for the nearest cafe, chatbot will propose a few points that are nearby to the client showing Viber maps and moving to Google maps. Also, chatbot presents scored points and helps to understand what value from it customers is having now. If a customer is satisfied or has the offer to improve services, there is a possibility to leave feedback.

The new approach for building chatbot features impressive customization and easy-to-use interface. So, Chatbots.Studio became a provider who helps increase cross sales and improves engagement with the existing customer.

Duration: 2 months.

Development team

Developers, Project Manager QA Engineer.

What we built
The solution includes

Viber Messenger

The services and Frameworks we used

We selected a set of tools matching our project requirements and client needs.

Viber REST API  + library messaging-api-viber, i18n were used for localization.

Winston and AWS cloudwatch  are monitoring and management services built for developers. In our case, we built logs using the aforementioned tools.

.env lets you customize your individual working environment variables and were used for storage of variables.

Eslint is pluggable and configurable linter tool for identifying and reporting on patterns in JavaScript. In our case, it was used for code formatting.

Aws EC2 is a web service that provides secure, resizable compute capacity in the cloud.

AWS CodeBuilt is a fully managed continuous integration service that compiles source code, runs tests, and produces software packages that are ready to deploy.

AWS CodeDeploy is a fully managed deployment service that automates software deployments to a variety of compute services such as Amazon EC2, AWS Fargate, AWS Lambda, and your on-premises servers.

Node.js + Express.js were used to build backend API for the bot that works in сhat.     

Amazon DynamoDB is a key-value and document database that delivers single-digit millisecond performance at any scale.

Development process

We know that a client wants to know how his/her product is developing, that’s why we gave sympathetic consideration and described in details each step of the development.

First development team chose tech stack. After this developers constructed architecture and development road. We made a huge roadmap and helped with the project all along with architecture decisions. Then created a basic authorization for users and implemented API methods that provide 2Cups. Development team built our own webhooks for the client needs.

The next step was full testing of the product and bug fixing. After that, we started the second version. Run it for all cafes in Ukraine.

The last step was to correct the auto-deploy system for the future upgrades of the product.

Bot architecture
The result

Chatbots.Studio team created chatbot for the cafe chain to optimize engagement with customers. The company has a 34% cross sales increasing after implementation discount and offers system into Viber via chatbot. Obviously, that new sales channel brought more revenue to the client and helps to build trusting relationships for B2C business.

Back to blog

Automation of the Claims Gathering for Insurance



Project description

We are building a multi-service system which will essentially allow users to fill out the insurance claims.

Insurance companies will be able to pass their clients over to our service to easily fill out the claim by themselves or with help of a company agent.

There are additional services used to make the claim filing more comfortable: medical appointment booking, number plate recognition, fetching vehicle data by its number plate, knowledge base (help service at any point of claim filing process), connection with live agent (like Intercom) inside the chat flow, Google Maps to easily set accident locations, address and more.

Duration: since June – ongoing

Development team
Developers, QA Engineer, Project Manager
What we built


The solution includes

Web widget allows users to create the claim.

SMS Twillio provides notification mailing to a user.

Administration part provides setting up the claims, new companies addition and manage the access. 

Alexa voice bot allows users to create the claim via voice.

Business Process Management System allows to discover, model, analyze, measure, improve, and optimize claims gathering processes.

The services and frameworks we used

BotKit is the open source platform was used to build bots with inbuilt Socket server.

OpenKB is a search based Knowledgebase (FAQ) backed by Lunr.js indexing to check created before the list of questions in Watson IBM. 

Watson IBM  is a question-answering computer system capable of answering questions posed in natural language.

Corezoid is the cloud OS that was used to track users’ activity and send a notification regarding incomplete claims.

Twilio is the cloud communication platform that was used for SMS notification regarding incomplete claims.

AWS Recognition  was used as a deep learning-based service for car’s number picture recognition.

Intercom is the messaging platform that provides online chatting insurance agents with clients.

Onfido is online platform that was used to verify people’s identities using a photo-based identity document.


Development process

The client sends the features which are discussed with the development team. After the detailed request, the developers map out a plan of development and start the working process.

Then the team starts implementation of frontend and backend parts of the bot.

When that part of the solution is ready we conduct internal and external testing.

Bot architecture

The result

Our team built the bot that gathers and passes processing clients’ insurance claims. It simplifies engagement between insurance representatives and the claimant. There is no need to fill out long and complex document paper – the bot makes it by itself.

There are additional services used to make the claim filing more comfortable: medical appointment booking, number plate recognition, fetching vehicle data by number plate, knowledge base (help service at any point of claim filing process), connection with live agent (like Intercom) inside the chat flow, Google Maps to easily set accident locations, address and more.

Get a Quote