Multimodal chatbots on the Telegram platform, orchestrated by a Large Language Model (LLM), fuse text, image and speech processing, filling the gap for natural multi-channel interaction.
Purpose. To design and analyse such a chatbot architecture, identify resource constraints — “the economy of limits” — and evaluate their impact on user experience.
Method. After a literature review (2023–2025) was created a prototype (Python + Telegram Bot API) based on GPT‑4-class LLM orchestrator, computervision, ASR/TTS and Retrieval-Augmented Generation modules. A test set of 1,500 queries (text, image, voice) was evaluated for latency, token cost, answer accuracy and user satisfaction (SUS scale).
Results. Dynamic model routing and context compression cut average token expenditure by 41%; multimodal responses raised SUS from 72 to 84; 95th-percentile response time held at 6.8 s. A hybrid knowledge store reduced hallucinations by 36%.
Conclusion. Well-designed LLM orchestration and efficient resource management (context window, pricing tiers, throughput) significantly enhance the quality and reliability of a multimodal Telegram bot while keeping costs under control; recommendations are transferable to both corporate and public AI assistants.

This research paper explores the capabilities and advantages of GigaCode, an AI-powered coding assistant developed by SberTech, for optimizing software development processes. This research provides a comprehensive analysis of GigaCode’s functionality, including its comparison with similar solutions such as GitHub Copilot and Amazon CodeWhisperer. Key aspects of the research are the next: Evaluating GigaCode’s impact on software development speed and code quality. Analyzing its integration capabilities with popular IDEs (VS Code, JetBrains, Android Studio) and cloud platforms. Providing practical implementation recommendations, including preliminary audits, pilot testing, and staff training. Research GigaCode’s unique features such as multilingual support, code generation from descriptions, refactoring, and automated documentation. The results of the research are confirm that GigaCode helps reduce development time, minimize errors, and enhance team collaboration. Special attention is given to its applying in education and business projects. The paper highlights GigaCode’s potential for further development and integration into corporate IT ecosystems, while noting the need for adaptation to address non-standard tasks.

The problem of automatic diagnostic of diseases, in particular cardiovascular diseases by ECG signal (Electrocardiography). Goal. Improve diagnostic efficiency by using modern machine learning methods. Results. The results of the article review are presented in a systematic form, the most effective methods and approaches are highlighted, as well as their applicability to solving specific problems of diagnosing cardiovascular diseases. Based on the analysis of existing works, conclusions are made about the prospects of using machine learning methods in the field of diagnosis of cardiovascular diseases based on ECG data and suggest possible directions for future research in this field. Practical significance. The article highlights modern methods and approaches used in modern research aimed at automatic detection of cardiovascular diseases using ECG data.

This research introduces a specialized Lacmus dataset designed for detecting missing persons in aerial photographs obtained from unmanned aerial vehicles (UAVs). The dataset comprises 1552 images with over 5000 annotated bounding boxes captured at five distinct locations characterized by grassy areas and sparse forests across different seasons. The primary objective of the study was to optimize the accuracy-performance ratio of YOLOv8 models based on the presented dataset. Experimental research has revealed that the best results were achieved using a medium-sized model with increased input image resolution without prior segmentation into smaller resolution images. The developed dataset and research results are intended for practical application in search and rescue operations, which could potentially enhance the efficiency of rescue missions and save human lives.

The article discover methods of organizing cyber attacks using artificial intelligence (AI) in the context of growing geopolitical tensions. The analysis of the use of AI by intruders for the subsequent development of protection systems is carried out. The applied method of analysis of MITRE and Lockheed Martin models, as well as considering real cases, make it possible to identify threat groups and techniques for using AI in cyber attacks. The authors substantiate the conclusion that it is necessary to create self-learning security systems against “combat AI”.

The topic of implement artificial intelligence into military activities is becoming extremely relevant in the context of the rapid development of technology and the changing nature of modern armed conflicts. Ignoring these trends is fraught with technological lag and a decrease in the state’s defense capability. Real-life examples of the use of intelligent systems in combat conditions indicate the beginning of a new era in military affairs. The purpose of the study is to comprehensively analyze the prospects for the use of artificial intelligence in the military activities of the Russian Federation and formulate recommendations for overcoming existing organizational, legal and personnel barriers. The scientific significance of the research lies in the systematization of existing approaches to the integration of AI into the military sphere, the formation of theoretical foundations for the adaptation of organizational structures and legal regulation. Practical significance is shown in the development of specific recommendations aimed at improving the effectiveness of defense structures through the use of artificial intelligence technologies. The main results and conclusions of the research work are the next. As a result of the study, measures were proposed to reform the organizational structure of the command and control of troops, develop human resources, form an adequate regulatory framework and expand the interaction of the army with the scientific and industrial community. Special attention is paid to the issues of safety and responsibility in the implementation of intelligent systems. The need for a multidisciplinary approach and constant adaptation of standards to technology development is emphasized. The scientific novelty of the work is the development of a comprehensive model for the integration of artificial intelligence into the military structures of the Russian Federation, taking into account organizational, personnel and legal aspects. For the first time, a phased introduction of AI technologies has been proposed, depending on the level of readiness of troops and infrastructure, as well as mechanisms to minimize the risks associated with the autonomy of intelligent systems. The research results can be used for the practical transformation of defense structures in the digital age.

The paper studies the features and prospects of artificial intelligence (AI) development. The authors made an analysis of the trends and challenges in this area and proposed the measures aimed at controlling AI impact on economy, society and state policy. The investigators also consider the issues of education, safety, ecology and regulatory policy that are necessary for full and safe implementation of AI-based technologies. Much attention is given to staff training, energy efficiency, data safety and ethical standards to use technologies. The research result has conclusion: it is necessary to design a complex interdisciplinary strategy which could integrate the interests of business, science and society to use efficiently AI and minimize possible risks.

Introduction. The development of services within Russia’s Unified Centralized Platform (UCP) in the social sector will focus on increasing the participation of retirees and self-employed individuals in managing their savings, as is done in several foreign countries. The possibility and conditions for creating a community of Pension Fund participants on the UCP platform are being studied with the aim of creating a community of collective investments to optimize additional income and reduce risks.
Objective. Based on international experience, it is proposed to develop a set of services on the Unified Centralized Platform of Russia in the social sphere, managed primarily by the Pension Fund, which will allow pensioners and self-employed citizens to invest their savings and receive additional income.
Methods. The methodology was based on analyzing foreign practices and methods of using pension savings, an assessment of the possibilities of the UCP and the subsequent development of a mathematical model. The rationale for this approach is to potentially provide tax benefits to individuals who save for retirement. For instance, contributions to the pension fund could be partially or fully tax-exempt, while investment income could be taxed at a reduced rate. This makes pension savings more profitable compared to bank deposits. To facilitate investments in various funds, services are offered access to educational resources to help potential investors better understand the market, assess risks, and make informed investment decisions.
Conclusions. It is proposed that pension funds, through the UCP, promote the creation of investment communities where citizens with common interests or goals can pool their resources to form a collective portfolio. This would diversify risks and increase profitability through a collaborative approach. For example, a community could invest in large-scale projects typically inaccessible to individual investors.
Results. An algorithm is proposed for creating a consolidated investment portfolio based on the Markowitz portfolios of participants (retirees or self-employed individuals) to reduce investment risks and optimize returns.