Cистема рекомендаций на сайте
KazanExpress logo

System for personalized recommendations on the Kazan Express website

Project information
Kazan express is a marketplace for Chinese and Russian good with free one day shipping
Has been working since 2017. The fastest growing marketplace in Europe.
Increase the number of purchases
KazanExpress.ru website users
Project concept
The customer sees recommended products that they may want to purchase.

Development process

Step 01
Step 02
Data preparation
Step 03
Architectural design
Step 06
User testing (KazanExpress employees)
Step 05
Train the model
Step 04

How does the system work

User visits client's site
Selects and product and places it in their basket
товар добавлен в корзину
The system offers products that the customer may need
с этими товарами покупают
Average check value rises

What the recommendations are based on

Purchase history of a specific user
Comparing purchases of different users
We look at the customer's «wish list» and previous
We do not take into account:
location and gender.
Two users have similar purchases - we call them «twin users».
We recommend what the «twin user» has already bought.


Spark library

framework for processing large data streams

ALS matrix decomposition

an algorithm that processes user actions in real time
intentionally limited the system for users who have made at least four purchases
Make bi-weekly system updates to prevent a system overload
A marketplace is a huge array of data (several million rows in tables).

That's why we:

Difficulties faced


Due to no clear quality criteria, acceptance of the project was delayed


It is necessary that during negotiations with the client that a clear understanding of the quality of the metercs and means of aquiring them is established


At times the service worked slower than we wanted


We found a hard-to-reproduce bug in the code and made additional optimizations to the speed of the algorithms.
«The only complaints we received during testing was that a man could have a pink phone case recommended, however we resolved that issue».
Project team
Nadezhda Zagvozkina
Data Scientist
Oleg Durygin
Junior Data Scientist
Александра Щетинина
What we have learned
Work more thoughtfully with recommender libraries
Handle huge amounts of data competently
Our plans
We are negotiating with the client for support and improvement:
new parameters (gender, age, location)
«We are planning to further develop in recommendations and build up competencies».
Where the implementation can be applied
Marketplaces, online stores and, in general, any entity that sells goods.
However recommender systems work best when given large data sets
Contact us