Danil Sidorov
RU EN DE
Telegram

Store with a recommendation neural network

An online store with a built-in neural network recommendation system

AI / E-commerce

Live interactive demo stand

An exact copy of the project interface and business logic with demo data. All actions run in real time.

Подбор похожих товаров на нейросети по истории и характеристикам

StackC#, .NET 5.0, ASP.NET Core, Unity, MS SQL Server

Safe perimeter (NDA): production logic on test data

What to test in the stand:
  1. Выберите товар в каталоге и изучите 10 нейросетевых сигналов
  2. Оставьте отзыв с оценкой для обновления матрицы предпочтений
  3. Запустите нейросетевой расчёт рекомендаций (Topology 10-4-1)
http://localhost:44381/api/
Live Stand
Recomendation System

Каталог товаров интернет-магазина Unity UI & SQL Server

Клиентское мобильное приложение на Unity 2021 обращается к REST-серверу на .NET 5.0. Список категорий строится динамически из таблицы [Type], а список товаров загружается по типам через /api/itemsbytype?type_id=....

База данных и категории продуктов 7 категорий • 14 товаров

Turnkey implementation

Need a similar system for your business processes?

I will develop the architecture, fault-tolerant backend, and fast interface. Protected by NDA, clean transferable code, and transparent milestones.

  1. Free timeline and budget estimate within one day
  2. Contract and invoice
  3. Warranty and post-launch support

On weekdays usually within a couple of hours

I will reply personally, no aggressive sales pitch

Task

Shoppers left product pages without seeing anything similar. The bestsellers block showed everyone the same items and missed their interests.

Solution

I built a store management app in .NET and connected a recommendation neural network. It suggests similar products based on purchase history and attributes. The model is separate from the interface, so it can be retrained on fresh sales without touching the main code.

What was done

  • Designed and built the architecture of the recommendation network
  • Replaced the generic bestsellers block with the neural network
  • Added recommendations of similar products by history and attributes
  • Separated the model from the interface to retrain it on fresh sales
  • Put recommendations and stock into one manager window in .NET

Result

Product pages now have a live block of items bought together with that product. The manager sees recommendations and stock in one window and adjusts the range from there.

Technologies

C#.NETT-SQL

Tell me about your task

Telegram