The key directions of the conference were the performance of AI systems and the reliability of IT systems. Experts from Sber, SberTech, Avito, Ozon, Alfa-Bank, Ecom.tech, Sovcombank, Positive Technologies, and other companies analyzed practical tasks: from load testing and monitoring to fault tolerance, working with large language models, and applying AI tools.
Special attention was paid to the application of artificial intelligence. At a round table moderated by Alexey Svetlov, MLOps engineer from IT1, CIOs and CTOs of Alfa-Capital, Gazprombank, and Flawway discussed strategies for implementing AI in Russian enterprise companies and how new technologies are changing the work and competencies of engineers. The focus was on AI agents, automation of load testing processes, and approaches to testing large language models.
Another important direction was the reliability of IT systems: participants analyzed real failure cases, methods for checking fault tolerance, and approaches to service monitoring. According to Alexander Krylov, CTO of RTK IT School and moderator of the round table, for engineers, the conference is an opportunity to compare their own scenarios with the practices of colleagues, and for managers, to understand what data the team needs to make decisions about service reliability.
Within the framework of the conference, Performance Lab presented updates to "Boomk" – a domestic solution for load testing. One of the key innovations was allocation groups – a tool for flexible load scaling. Now "Boomk" allows running a single JMeter test plan, distributed among several generators, with unified metric aggregation, or simultaneously using different test plans on different machines. For the On-Premise version, support for the domestic technology stack has been expanded: the solution works in conjunction with Astra Linux OS and Tantor SE DBMS. The platform's AI capabilities are also being developed: comparative AI reports, the ability to set an analysis focus, and the import of external metrics from third-party monitoring systems have appeared.
Another significant event was the presentation of the Russia Quality Report 2025–2026, a unique study of the Russian QA market, which Performance Lab conducts every two years. The issue examines the impact of AI on workflows, the connection of quality with business metrics, and the requirements for modern engineers.
A significant part of the program was dedicated to engineering cases. Alexey Kaznacheev, Project Portfolio Management Director of the T1 IT holding, spoke about the transition from an engineer's role to a team leader's role, the value of systemic engineering thinking in management, and the role of engineering culture. Ksenia Voronchagina, Deputy Technical Director of Performance Lab, analyzed the features of load testing 1C systems, technical nuances, and ways to improve process efficiency. Ivan Prikhodko from Ozon spoke about the features of load testing in the company and focused on the use of IP spoofing in tests.
According to Vladimir Kashirsky, CEO of Performance Lab, this year's conference focused on AI, language model performance, and service reliability. At the same time, it was important for the organizers that participants could not only get acquainted with the solutions of colleagues but also understand their limitations, find ideas for their own projects, and new professional contacts. Alexander Makarov, Technical Director of Performance Lab, noted that the development of the industry is clearly visible in new approaches to fault tolerance and performance, AI-based tools, and constantly evolving testing practices.
iXBT acted as the information partner of the conference.