On April 30, 2026, Ruslan Safarov, a research lecturer at the Department of Chemistry at the L.N. Gumilyov ENU, a postdoctoral fellow at the Dosmukhamedov Atyrau University, held an online seminar for university staff as part of the Science CAFE program.

The topic of the seminar "AI-Powered Researcher: The full cycle of preparation of the Q1 article" was devoted to the practical application of AI tools in scientific work. The speaker began with a key thesis: AI does not write an article for a researcher – it removes the routine so that a scientist can do science.

During the seminar, the problem of the traditional approach to writing articles was analyzed: the introduction is written blindly without ready-made results, the literary review is duplicated twice, and routine tasks accumulate in one exhausting "sprint". Instead, the speaker suggested the reverse order of writing: first, the Results and Methods, then the Discussion and Introduction, and only at the end – the Abstract and the title.

Special attention was paid to the researcher's AI ecosystem, which includes six interrelated tools.: Zotero for bibliography management, NotebookLM as an isolated knowledge base without hallucinations, Python for data analysis and visualization, MCP for offline agent access to scientific databases, Google Workspace and Obsidian for task management and personal knowledge base. At the center of the ecosystem is a research agent (Google Antigravity/Claude Code) that reads data, calls tools, and automates routine sections of an article.

Using specific examples, it was shown how the search for 50 articles is reduced from several days to 3 minutes, data analysis from CSV turns into a publication graph with a descriptive paragraph, and sections such as Author Contributions or Cover Letter are formed in 30 seconds per prompt.

The seminar was held in the format of a Zoom conference and aroused great interest among the audience seeking to increase publication activity in highly rated international journals.