Artificial Intelligence in the Company

Artificial Intelligence as a Response to Resource Shortages

The implementation of artificial intelligence in corporate processes is no longer just a technological trend but a pressing necessity, especially amid staff shortages. The presentation details the experience of MegaFon, where a team of 10 HSE specialists serves over 22,000 employees nationwide, including branches, communication stores, and subsidiaries. In such a situation, automating routine tasks is the only way to maintain efficiency and provide 24/7 personnel support.

From Voice Assistants to Corporate Chatbots

The speaker demonstrates how the company, already experienced in creating client voice assistants (such as Eva), scaled these technologies to internal processes. Prompters and virtual mentors are used for employee onboarding and training, helping newcomers master customer service standards. A logical continuation was the creation of the internal chatbot "MegaColleague," which took over first-line support functions for a wide range of issues, including HSE.

How "MegaColleague" Works in HSE

The chatbot operates on Telegram and is built on the principle of a multi-level response:

• Brief and to-the-point answer: AI generates a direct response to the user's query, allowing for quick problem resolution without unnecessary information.
• Navigation through instructions: The bot provides links to relevant internal documents and regulations (e.g., first aid kit usage rules or PPE issuance procedures).
• Clarification of nuances: The system explains what can and cannot be done in a specific situation, reducing the risk of violations.

This approach, as the speaker notes, avoids additional clarifying questions that often arise with overly concise answers.

Implementation Stages and the Role of the Business Customer

The chatbot creation project took only three months. A key success factor was the close collaboration of business units (including the HSE department) with the Data Science team. The process included:

• Knowledge base formation: Preparation of high-quality instructions and materials for model training.
• Manual testing: Initially, specialists analyzed hundreds of questions and answers daily, adjusting algorithms to ensure accuracy and legal compliance.
• Automation and analytics: After launch, the system switched to auto-test mode, and manual intervention was minimized.

Results and Development Prospects

The implementation of "MegaColleague" consolidated consulting channels and reduced response waiting times from several days to seconds. This not only increased employee satisfaction but also freed up a significant amount of working time for HSE specialists (estimated at thousands of hours per year), allowing them to focus on strategic business tasks.

The company plans to transition from informational support to proactive actions (agents), where the chatbot can independently perform operations, such as placing a PPE order considering the employee's individual parameters.

What you will learn from this webinar:

• How to compensate for the shortage of HSE specialists using AI?
• What is the optimal response structure of a corporate chatbot to employee questions?
• How long does it take to develop and train an internal AI consultant?
• How to organize the chatbot testing process using business units?
• What prospects does the personalization of AI services in HSE open up?