Our client faced a number of challenges related to the use of AI, including the fact that employees had to register with separate AI services to perform different tasks, as not all LLM platforms are equally effective (for example, Claude is better for coding, while ChatGPT is better for text generation). The company’s internal AI projects also required overseas hosting.

In response to these challenges, we implemented a unified tool that allows company employees to work with all the necessary AI models. The solution is based on open-source developments, offers rich functionality, and can be integrated quickly.

We deployed and configured the infrastructure, connected it to the API, and grouped users to grant them access based on business roles. For example, specific models are provided for system and business analysts that are best suited for solving their tasks.

In addition, we implemented a usage statistics collection system. Analytics generated monthly show how frequently and to what extent employees use the AI models. The collected data influences budget forecasting: payment for AI usage depends on the number of tokens (words) that users request from the model and receive in response. There is also the possibility to track that LLMs are used for work-related matters, not for personal needs.

We are now continuing to update the product and keep the AI models (ChatGPT, Gemini, Deepseek, Mistral, Claude) current, following the latest trends and technology updates.

The implementation of our tool helped the client not only solve the issues with AI access but also create a more productive work environment for its employees.

• Unified Solution. Employees log in to a single system and gain seamless access to up-to-date AI models, which simplifies the work process. 
• Managed Access. Group-based access configuration allowed the tool to be tailored to specific roles, increasing overall efficiency. 
• In-Depth Analytics. Regular analysis of AI model usage helps company management better understand employees’ needs and plan the budget. 
• Developed LLM Usage Culture. Previously, only certain specialists used them, but now 80% of employees access them at least once a week, making work more effective.