How to bring perfect order to a huge item catalog and masses of unstructured data

About the company

Rudnik Aprelkovo JSC has been mining natural resources across the country for 45 years.  It mines ores and placer sands of precious metals (gold, silver, and platinum group metals) and is one of the country’s largest precious metal mining companies. It is a large company with multiple divisions (a holding company).

Project goals

• Analyze the existing 1C:UPP-based system in order to correctly export the data, normalize it, and then load it into 1C:MDM
• Design the logic of a centralized entry process for all data types involved in the automation. Develop a step-by-step data entry system with clear areas of responsibility for specific users
• Introduce a data completeness and quality control system
• Automate the entry of standard items by creating new items as copies of existing ones with minimal editing
• Use an advanced search system (several mechanisms at once) to reduce the risk of duplicates and redundant entries

The most pressing task was to automate uniform data collection.
1C:MDM was chosen because the client already had 1C systems in place, so it made sense to add it as a natural extension.

For 15 years, items and counterparties had been recorded in a custom-built accounting system based on a heavily outdated 1C:UPP. Data from all the companies accumulated without any structure. The same object could appear under different names. Sensitive data could not be processed correctly. Many subsidiaries came into the group with their own accounting and automation.
We needed to bring order to the catalogs, introduce clear data entry rules, reduce the risk of duplicates, and implement master data expert functions.

We designed and automated business processes for populating data and controlling its quality.

A distinctive feature of the 1C:MDM implementation in this project is that items are classified not only by the main classification but also by several additional ones. Items in 1C:MDM are tracked across multiple dimensions (both management and accounting). For example, the same item may fall into different groups in accounting depending on its origin (purchased or produced in-house). The goal was to eliminate duplicate items while preserving a complex structure of relationships.
The system includes controls for counterparty uniqueness, as well as the built-in 1C:SPARK service for monitoring business ratings and reliability indicators.

During the project, we normalized more than 90,000 (!!!) primary records, cutting their number by more than half. We also cleaned up the item catalog, which contained more than 16,000 entries.

Together with the client, we designed and configured business processes in 1C:MDM for sequential data entry by employees, each responsible for their own block of data (who fills in the catalog item forms, how, in what order, and with what areas of responsibility). For better automation and a more intuitive experience, additional restrictions by roles and user groups were added to the system. The master data expert checked the item catalog for similar entries and, when necessary, added a new record, monitoring the completeness and correctness of the data. Fields that are critical to fill in were made mandatory at the system level.

Results

More than 20 functional gaps were closed. The new system met all of the client’s needs. 

• A system linking the main item classifier to multiple additional item accounting dimensions was developed and configured.
  New processes were introduced that made work efficient and accurate.
   Employees work faster
• The quality of data in the system and the quality of accounting and reporting improved.
• Document generation became simpler. 
• A master data expert panel was introduced to control data entry. 
• A centralized data entry system was established, from which data is distributed to all of the holding’s databases and systems.

Benefits the company gained from the project

• The company can find the right items and counterparties in catalogs quickly and accurately. Expert validation of the data helps avoid errors and inaccuracies.
• Data accuracy has improved, which in turn can improve decision-making and help the company plan its operations better.
• Automating some data entry processes reduces the likelihood of errors that could lead to additional costs.
• Putting items and counterparties in order gives the company a more structured and organized database. This makes it easier to find and access the information it needs and simplifies data analysis and data-driven decision-making.
• Efficiency, data accuracy, and data quality have increased, costs have decreased, and data management and customer experience have improved.

Ultimately, 1C:MDM significantly improves data quality not only in the original system the data came from, but also in all the systems to which 1C:MDM distributes high-quality information.