Software development partnership with PrettyLittleThing - a leading fashion retailer in the UK

Founded in 2012, PrettyLittleThing grew very fast and became of the leading fashion retailers in the UK. Therefore, the company needed to expand its development capabilities fast to scale their software products, add new functionality, automate some processes, and ensure high performance and security of their solutions. All this was necessary to better meet the needs of their growing customer base. At the outset of our cooperation, we got the following tasks:

• extend the in-house team with strong software engineers experienced in PHP, various JavaScript frameworks, AWS, and other technologies;
• automate the process of product creation on the client’s e-commerce website that has over 30,000 items;
• speed up the software development process.

• Establishing effective communication and alignment among different distributed teams;
• Developing the functionality that helps easily manage thousands of products.

IT specialists at N-iX belong to two different teams of the client. They include software engineers who have experience with PHP and JavaScript, a team lead, a manual QA specialist, and a project manager. Together with PrettyLittleThing, we work on the back-end of the website. We help develop custom functionality for Akeneo, a product information management system. Also, we work on shipment, order return, and re-order functionality implemented with the help of AWS Lambda.

PrettyLittleThing leverages Akeneo to automatically create products with different attributes. The system is utilized by the procurement team. They use Microsoft Excel to create a list of products that need to be added. N-iX developers have created functionality that allows uploading Excel sheets to Akeneo. We apply the ETL process for processing data. When a new Excel is uploaded to the system, a PHP сron parses it. Then, the import script converts it to an array of data, goes through each row of the array, and selects only those rows that are relevant. It converts the data into codes, accommodating prices for the particular region, and automatically creates product models. Then, new products are created.

Also, we have developed custom functionality that ensures the transmission of this data further to Amazon SQS, where other services such as warehouse management service, order management service, discount management service, inventory service, and others can get this information. Each of them has an independent database that displays the most recent changes. So if one of these services stops working, the rest are up and running.

Besides, we have built a new index in Elastic Search, a NoSQL database which facilitates the search of products. Our engineers developed the functionality that lets them search products by the category name

Also, we are working with the client’s team on building lambda functions on AWS with Node.js. AWS Lambda along with Amazon API Gateway allow implementing serverless REST APIs on the AWS platform. To do this, we configure REST API in Amazon API Gateway and map individual API operations to specific Lambda functions executed with Node.js. The Lambda function interacts with the database and runs whatever logic is needed to process the request.

As a part of our services, we provide professional manual testing that includes iterative and regression testing. Also, our developers write unit tests with PHP and JS to test particular segments of code.