AI Evaluates the Quality of Employee Briefings at IrkAZ
The system has been in operation at the plant for a month.
The Irkutsk Aluminum Plant has introduced a system that assesses the quality of employee briefings. It uses two neural networks. The first records conversations between managers and employees and converts the audio into text. The second analyzes the transcript to identify the strengths and weaknesses of the briefing. The system then recommends ways to improve the process.
“For the first time, we’ve developed a tool within the company that provides objective, immediate feedback. Previously, we could assess speaking quality and how thoroughly a briefing covered its material only on a case-by-case basis. Now, AI analyzes every shift and helps supervisors improve their public-speaking skills. Most importantly, it makes sure critical occupational safety information reaches every employee,” said Andrey Grushevsky, digitalization manager at IrkAZ.
The system has been running at the plant for a month and now covers every department. It is no longer possible to miss a briefing. The project was developed with support from RUSAL’s Digital Solutions Department.
