App Scanner

The client wanted to develop a tool that eliminates the risk of human error on investment data entry and reduces the number of the errors they get on throughput.

Industry FinTech

Duration: 3877 hours

Team: 8

Download PDF

Technologies utilized

Python, .NET, MySQL, Tornado, App Engine

Team composition

1 Team lead
1 Tech lead
1 Solution architect
1 Project manager
2 back-end developers
1 Manual QA specialist
1 Automation QA specialist

Have a similar project?

Estimate

CLIENT

Octopus Lab is the fintech innovation unit of the Octopus Group. They build smart financial products and use technology to drive cultural and process change, improving customer experiences and value for money.

Challenge

The client needed a way of automatically processing that forms data with information on investments and integration this into other systems they use [Novus and Watson].

Features

Project Management

The project was delivered using a Scrum methodology with bi-weekly sprints, where we worked with the client on requirements elicitation to define a maintained backlog of business needs.

Modifications

We have changed key services without regression for other services and made modifications to the structure of the database. Added algorithm for digitizing images and made the REST API for key services.

API

One of the key challenges was the integration of the systems through the new API, which was solved by creating an “Orchestrator” for the other systems with their own logic to manage the data traffic between them.

Business values

Modernization

The API and Orchestrator that were built as part of the solution for the client, applied process validation rules to check the quality of the data being passed through the system.

Convertion

The solution allowed users to convert images into a text representation (Optical Character Recognition - OCR). When the user downloads an image or a set of pictures on the server, then which text recognition, removal of unnecessary elements, grouping of text in blocks/paragraphs, and saving for further editing occurs. At the output of the service, a structured text is obtained, which was found in a certain zone/zones in the image.