HSE as the Spark of Digital Transformation: How We Launched AI in the Company

Digital transformation in large industrial companies often faces conservatism, bureaucracy, and a fear of innovation. Health, Safety, and Environment (HSE) specialists are forced to spend time on routine processes instead of focusing on preventive measures. In her presentation, Anastasia Lukovenko analyzes a practical case of overcoming these barriers through the phased implementation of artificial intelligence tools.

From Simple Chatbots to Management Engagement

The integration of neural networks does not necessarily require massive IT budgets at the start. The speaker demonstrates, using the example of her company (over 18,000 employees), how the initiative began with basic solutions created without involving programmers. Specialized chatbots were developed:

• Risk assessment bot and safety quizzes — to engage personnel and clearly demonstrate the technology's capabilities to management.
• Information assistant for PPE and Special Assessment of Working Conditions (SAWC) — a tool that allowed workers to quickly find out their workwear issuance rates and working condition classes. This radically reduced the time HSE specialists spent searching for data in local regulations.

The success of these pilot projects led to scaling: from local master classes for HSE specialists, the company moved to targeted training for top managers and holding strategic sessions on implementing AI in other departments.

Information Security and In-house LLM

As expected, information security became the main barrier to AI development in the corporate sector. The use of external platforms raises justified questions from the security service about server locations and data leak risks. The presentation details the solution to this situation: transitioning from cloud solutions to deploying an in-house local platform. Purchasing servers and installing a local large language model (LLM) allowed the company to start training the neural network on internal documents while remaining entirely within a closed security perimeter.

What you will learn from this webinar:

• How to build a chatbot to consult employees on PPE and SAWC without programming skills?
• How to overcome top management skepticism and justify the need for investments in AI technologies?
• What arguments and technical solutions will help pass approval with the corporate security service?
• How to organize the transition from public AI services to an in-house local LLM on the company's servers?