Which technologies we work with in AIS Novations

Tatyana Koneva
AIS Novations
Published in
4 min readJun 28, 2019

We use various modern technologies for in the development process. As new solutions and instruments are released, we try to test them and introduce them in our projects. Here we describe several real cases from our experience.

Dynamo

A big online remortgage platform for British market. It give user an opportunity to take a credit on a preferential terms for a continuous period, in order to reduce monthly payments.

Time: Aug 2017 — ongoing

Team: 5 developers, 2 QA

The customer wanted

An automated remortgage platform.

The technologies used

  • Node.js
  • MongoDB
  • SCSS
  • HTML
  • JS
  • React
  • Redux
  • Redux-Saga
  • BrowserStack
  • InVision
  • Jira
  • Confluence
  • Swagger
  • Slack
  • TestTrail
  • Onfido (IDV)
  • MailChimp
  • Mandrill
  • Zendesk
  • Azure Storage

How we solved the task

We have launched an MVP in 3 months. The project was divided in 3 portals: Website, Customer Portal, and Middle Office, that were then united into a big efficient system.

The remortgage process was automated. The support service was designed: 500 of consultants are ready to help clients in a chat or via email. The platform was synchronized with external systems, such as Mandrill, Mailchimp, Zendesk, Onfido. Fixed mortgage rates and a payments calculator were added.

Kwick Doctor

A service that remotely connects patients with doctors. A User can make a video-call with of the available doctors, get a paid consultation within 10 minutes, request a drug order, or a sick note.

Time: December 2017 — ongoing

Team: 2 developers, 1 PM, 1 BA

What the customer wanted

A service with admin-panel for remote connection between patients and doctors.

The technologies used

  • React
  • Redux
  • Redux-Saga
  • WebSockets
  • TokBox (WebRTC)
  • REST API
  • React Native
  • Elixir
  • Phoenix
  • PostgreSQL
  • Redux-Router
  • SCSS
  • Style-in-JS
  • Postman
  • Storybook
  • Jira
  • AWS S3
  • Jenkins

How we solved the task

We have created the layouts, worked out a specification and moved to the platform architecture stage. There were created the Mobile and SPA applications: web-interface for doctors, and mobile version for patients. SPA was also designed with admin interface.

Messaging, video- and audio calls were automated. The prototypes for doctors’ flow-onboarding were prepared. Supporting three repos to avoid the app functionality problems was not the easiest task.

Helpful

A start-up company from London. Designed to get rewards for the plastic bottles recycling. A User is supposed to scan a logo on a bottle or a plastic cup using one’s smartphone camera. The application recognizes it and gives the information about the possible ways of recycling this bottle/plastic cup. Then the application builds a route to the nearest plastic collection point.

After the bottle/cup is given to the collection point, the user gets a HELPFUL-token, that can be then used for buying the eco-products designed to reduce the plastic consumption.

Time: November 2018 — ongoing

Team: 2 developers, 1 PM

The customer wanted

An Android and iOS application and a website using Azure Custom Vision Service for image recognition.

The technologies used

  • Azure WebApp
  • Azure database
  • Azure custom vision
  • Ionic
  • Angular 6
  • Apache Cordova
  • Asp.net core
  • MS SQL
  • ADO.NET Entity Framework

How we solved the task

We have built the project using Apache Cordova for both Android and iOS. While working with Cordova, several issues regarding the compatibility, the absence of support, and the lack of plugin functionality have occurred. As a result, we have edited the plugins (built natively for both platforms).

We needed to implement “live search” of an object within the frame. The API for objects recognition didn’t fit because of the poor speed and the high price. That’s why we were using TensorFlow-models, that are launched on a user’s device, and work faster than the API. In order to make the application scalable, we were using a cloud platform Azure.

--

--