● Implementation of technical expertise and leadership in the field of GenAI-based solution design within the framework of the RnD team for the development of SberOS, including overseeing the implementation of the AIOS scientific approach as part of the deployment of self-hosted SLM on user servers. ● Solo Vibe ML/AI-dashboard coding with code-review and service prototyping on the Django/Streamlit stack (GigaChat API + LangChain/LangGraph +SciPy ecosystem) for analyzing CX + human digital twin as part of the production process of the SberOS product in order to integrate into the AI-native infrastructure of the Bank. ● Coordination of internal and external corporate communications between various departments of Sberbank and external structures. ● Participation in research on market trends in the field of agency and AI autonomy and the competitive advantages of innovative technical solutions in the field of LLM and LVLM development to increase the consumer value of the Bank's products. ● Work on improving the user journey and creating an effective roadmap for the development of products and services based on the corporate operating system, work with the pains of users of the corporate product. ● Conducting regular monitoring of compliance of the created solutions with information security and cybersecurity standards, including the code of ethics of AI and its security.

● Research in the scientific direction of Spiking NN: building custom neural network architectures based on torch.nn and snnTorch (LSTM with Attention, Spiking ResNet, Spike-Driven Transformer) with the support of LLM, conducting machine learning to obtain time sequence analysis models, work on compiling a dataset, inferring the obtained models using ONNX Runtime, optimizing the execution code. ● Development of exchange analytics and pricing projects using ML-pipelines (predictive time series analysis, ensembling classical approaches, writing trading bots and trader assistants using RAG, developing multi-agent systems based on local deployment of DeepSeek-R1 and Phi-4 model families, as well as API access to well-known providers) in Bitrix24 in different teams. ● Development of the concept of a smart trading platform through consultations in the structure of machine-building enterprises based on predictive analytics technologies within the framework of ongoing automation and digitalization in the production centers of Yekaterinburg. ● Decomposing and maintaining a backlog of AI tasks in medicine (personal assistant, software for personal gadgets and medical devices, datasets) at Atlassian Trello in various teams as a technology leader within the framework of the "Archipelago" accelerator conducted by "NTI". ● Help with the project of indie games on Russian literature on the Ren'Py engine using STT and Summarization ML-pipelines. ● Authorship of the modern educational course "Linear Algebra" in Python on the "GeekBrains" platform. ● Development and implementation of the course "Fundamentals of automated software testing" for machine-building enterprises of the Sverdlovsk region.

● Supervising the direction of the hub architecture of the central monitoring station, monitoring the development and deployment and revision of the code base of all versions of software and hardware gateways as suppliers of station data in the telematics product development sector, integration with the AI system, analysis of standards for software quality, system design from scratch. ● Leader of AI competencies and organizer of the RnD department, advising representatives of the project office on artificial intelligence technologies, researching the market of artificial intelligence systems in Russia and the world. ● Collecting requirements from end users (doctors, healthcare managers, junior medical staff): conducting interviews, descriptions of Use Cases, User Story, JTBD (knowledge and experience of describing the Customer Journey Map in medical institutions). ● Analyzing the prospects and capabilities of Russian single-board computers based on the ARM hardware platform, drafting technical specifications for the development and purchase of single-board computers with various technical characteristics, testing systems using various assemblies based on the GNU/Linux software kernel. ● Solving machine learning problems based on compiled models of convolutional neural networks and experimenting with SOTA architectures in classification, object detection and instance segmentation tasks using the PyTorch framework, implementing PoC, conducting and evaluating A/B test results, developing documentation for certification events, studying standards in the field of AI, setting tasks and evaluating cost and quality of dataset collection and data markup for CV systems, participation in thematic conferences, study of scientific literature. ● MLOps of ready-made telemedicine software solutions for Docker, including taking into account the orchestration of Docker Compose, on Windows and Linux.

Was serving as a technical expert in the process of building data-driven processes at "UOMZ" Production Association, paving the way for AI-driven infrastructure: ● Development and implementation of digitalization techniques based on the functionality of automated optical production: examination of the business model of the division, decomposition of tasks of transition from automation to digitalization within the framework of building a digital twin of the enterprise, formalization of restructured business processes in BPMN notation, presentation to management of the AI assistant architecture of the assembly line for integration into MES based on products of the Hadoop ecosystem (Kafka, HDFS, HIVE, and Spark). ● Presentation to the management of an architectural solution by an AI assistant design engineer based on AI technologies using the DeepPavlov framework in collaboration with NIC "Schwabe at MIPT", study of scientific papers on GPT neural network architecture. ● Analysis and elaboration of CASE tools for creating hologram projection systems for AR solutions in Python using the genetic algorithms of the DEAP library, launching a business process at the holding level with a technical justification for the development of AR glasses, teamwork in Siemens Teamcenter Engineering. ● Development of integration of modular computer vision systems on the Nvidia hardware and software platform using Jetson SDK Manager, construction of pipeline machine learning models in Python (image classification, object detection, diffusion), optimization of model execution, curation of dataset assembly.

As a certified research teacher, was engaged in: ● Leading research teams in the framework of project activities in the information system of the Federal Academy of Sciences of Russia, accompanying the educational trajectories of UrFU students in the Laboratory of Magnetic Semiconductors of the Department of Nanospintronics. ● Development of the strain-magneto-optics concept in the team of the nanospintronics department: processing experimental data using classical ML methods, elaboration and discussion of mathematical models of target scientific fields. ● Conducting research on the creation of a human spinal cord multiimplant for neurosurgical operations in cooperation with "Orfo". ● Organization of an interlaboratory youth group to discuss the hybrid hardware architecture of neuromorphic computing modules and their software instructions based on the Department of Nanospintronics.

Was conducting research on the human brain together with members of the clinic staff led by Doctor of Medical Sciences, Professor A.A. Belkin. I was engaged in the role of a medical physicist (Junior Data Scientist): ● Conducting data analytics on Statistica 6.0 and MS Excel in the Department of Functional Diagnostics and Neuromonitoring, processing data in specialized application software for managing measurement modes of medical equipment for conducting examinations of clinic clients and making diagnoses by somnologists, epileptologists, neurologists. ● Conducting medical research in collaboration with specialists from the neurosurgical department of GKB №40 using "Neurosoft" neurodiagnostic programs and the "Magstim" transcranial magnetic stimulation software and hardware complex (human brain mapping procedure), organizing the storage and processing of research data. ● Testing basic telemedicine practices and technologies based on the network infrastructure of medical centers in the Sverdlovsk region in the process of implementing HL7 standards and discussing the prospects and opportunities for building DWH in cooperation with the Ministry of Health of the Russian Federation.

As an employee of the Department of Magnetism and Magnetic Nanomaterials under the supervision of the Head of the Department, Doctor of Physico-Mathematical Sciences, Professor Vaskovsky V.O. and a visiting fellow of the Department of Medical Physics, Computer Science and Mathematics under the supervision of the Head of the Department, Doctor of Biological Sciences, Professor Blyakhman F.A. at the UrFU Institute of Natural Sciences and Mathematics, was engaged in the following: ● Leading the project activities of the UrFU Research Institute of Physics and Applied Mathematics in the areas of "magnetism" (nanospintronics structures for advanced data storage facilities, components for wearable gadgets), "metrology" (development of R&D efficiency metrics in universities) and "biophysics" (classical machine learning algorithms based on the OpenCV library for analyzing medical images in asynchronous tasks in the work of the human myocardium), the organization of interaction between two scientific schools (two departments of the Faculty of Physics). ● Organization of the uAnalytiCon circle and conference on analytical Philosophy in cooperation with specialists from the UrFU Department of Philosophy (Department of Ontology and Theory of Knowledge): discussion of the problems of cognitive modeling and prospects for the development of AI. ● Support of customers of high-tech products produced by legal entities associated with the UrFU Research Institute of Physics and Applied Mathematics at the stage of system analysis of business requirements.