Inevils has developed an intelligent ecosystem for DMT DVRs with support for live broadcasts, GPS and neural networks

2025: Building an intelligent ecosystem with support for live broadcasts, GPS and neural networks

The IT company Inevils has developed a large-scale software ecosystem for the Russian manufacturer of wearable DVRs DMT. The software was a response to a market request: previous decisions did not meet the requirements of users - an overloaded interface and opaque access to data slowed down the work of customers, including from the public sector. Inevils announced this on May 7, 2025.

The company decided to finalize the system, better adapt it to customer requirements. The platform was supposed to provide stability, flexibility of access and the ability to integrate with internal systems.

The development included the creation of a secure server part, storage modules and quick data search for many parameters, a tree access rights system, support for GPS and online broadcasts from DVRs. One of the key elements is an AI module trained to recognize safety violations: the system records the absence of a helmet, mask or gloves and automatically notifies those responsible.

For implementation, C/C + +, PyTorch, Django , PostgreSQL, QT and other tools adapted for Russian OS were used. The project was conducted according to the Agile and Scrum methodologies, with a focus on extensibility and reliability.

The software continues to develop: new functions are added, neural networks are updated, the set of scripts is expanding. The platform is designed for long-term use and is integrated into the DMT product strategy.

When we decided to optimize, we needed a partner who not only understands the task, but knows the specifics from the inside. Before the founding of Inevils, Dmitry Dudnikov was responsible for development in our company for several years. This cooperation switched to a new format, but retained the main thing - reliability, mutual understanding and the ability to quickly adapt to real requests. It was important for us to get a solution that can be developed within our own food ecosystem. The main tasks concerned stable operation, ease of use and the ability to control access to data at the organization level. Now the platform is solving these problems, and we continue to work on the next stages, said Andrey Gusev, Deputy Director of DMT LLC.

We initially understood that the question is not a one-time development, but the creation of an environment to which more and more requirements will be presented by users, customers and regulators. Therefore, the architecture was built with an eye to scaling, AI support and long-term operation in real conditions, noted Dmitry Dudnikov, CEO of the IT company Inevils.