Only RK

Price: 0

Number of applications: 2

Decision acceptance deadline

21.10.25 (inclusive)

Form of award

Under discussion

Product status

Idea

Task type

ICT tasks

Сфера применения

Media sphere

Область задачи

Information processing and transformation

Type of product

Software/ IS

Problem description

In medical databases and pharmacy catalogs, the same drug may have different spelling options (abbreviations, registers, spaces, release forms, and dosages). The task is to create an algorithm that: correctly recognizes and normalizes such variants.; groups them by identical drugs; returns a unified, "canonical" name; generates a data structure that is convenient for training an AI agent.

Expected effect

SKU Name Normalization System

Full name of responsible person

Severgin Anton Olegovich

Purpose and description of task (project)

To develop a system for normalizing the names of SKUs (medicines) to improve the accuracy and speed of searching through an internal database and external reference books, as well as for further training of the AI agent.

Note