Sibur has implemented an AI agent to automate the technical review of supplier proposals. The system compares submitted documents against technical requirements, identifies discrepancies, and automatically generates requests for missing information. This process has accelerated the initial technical evaluation of proposals by seven times.
The company manages a large procurement system, conducting approximately 8,000 procedures annually with over 2,000 suppliers, 85% of which are domestic. Before automation, up to 80% of the technical analysis time was spent on manual data collection, such as requesting product passports or confirming supply experience.
The AI agent retrieves technical data from the company's electronic trading platform and creates structured request templates for suppliers. This approach reduces manual document processing, minimizes human error, and speeds up communication. Suppliers receive clearer feedback, which helps in gathering complete documentation faster.
Beyond this specific agent, Sibur utilizes AI for searching alternative manufacturers, optimizing logistics routes, and identifying product analogs. The company has been applying AI technologies since 2018, with over 30 implemented cases and more than 200 initiatives currently in development across various business processes.
