Ovanes Petrosian
Summary: ex-Director of AI-center of St. Petersburg State University, which, in 2024, became one of biggest AI-centers in Russia. Professor of Department of Mathematical Modelling of Energetic Systems of St. Petersburg State University with vast 11+ years background in AI and commercial companies R&D, wide experience in technology, author of more 90 research papers, Dr.Sc. at 29, Professor at 30 years old, 1st places in world AI competition. Husband and recent father.
Information systems development interests:
·LLM-based personalization and software optimization: OS, app.
·Industrial controllers: software and controllers.
·Digital twins for industry: for industrial and individual engineering facilities.
·Creation of optimization solvers: solvers for MILP, MINLP, Black-box problems, AI solvers.
·Predictive analytics systems: predictive and proactive analytics.
·Schedulers for operations: mobile / cloud projects, industry operations.
Research / algorithm development interests:
·Optimization: Neural network training algorithms (including BNN training), Global optimization, Black-box optimization, Mathematical optimization (MILP, MINLP solvers).
·Scheduling: Offline and online scheduling using optimal / suboptimal / real-time algorithms.
·Optimal Control: Adaptive control, Model Predictive Control, Filtering (Kalman Filters and others).
·Forecasting and Anomaly Detection: Classical forecasting (regression models and others), Deep learning (LSTM and others), Boosting algorithms (LightGBM, Xgboost), Multidimensional forecasting methods, Anomaly detection methods.
·Explainable AI: SHAP, application of XAI for different AI algorithms.
·Multi-agent control.
·Graph computing: Distributed graph scheduling, online and offline graph scheduling.
·Operations Research: Energy and Telecom optimization control systems.
Application areas:
·Mobile / cloud computing, OS.
·Fundamental AI (new methods, architecture, approaches).
·Energy.
·Telecommunications.
·Transport and routing.
·Engineering and mechanics.
Work Experience
2022 – 2024: Director of AI-center of St. Petersburg State University.
From 2022 to 2023, I was responsible for developing the strategy of the SPbU AI-center, forming the RnD teams and inviting new industry partners. Bythe end of 2023, the new AI-center team applied and won the biggest Russia research and industry project for AI-centers in Russian. Under my management, we became one of biggest AI-centers in Russia with a team of more than 75 RnD engineers and developers working on 6 end-to-end AI products and RnD projects for different industrial partners.
Responsibilities: Designating AI Center strategic planning, managing partner’s collaboration and contractual obligations; participating in public speaking events, interviewing and hiring new researchers, and directing AI technology and architecture review, products idea generation, project cost estimations.
Achievements:
·2024 – Hired and organized a work of 6 AI teams for 6 AI end-to-end products:
oAIoT platform. The goal of the development is to create an effective platform for monitoring and managing IoT devices. The platform will combine IoT devices with artificial intelligence capabilities to provide intelligent data analysis, automation and decision making.
oAI algorithm optimization and compression platform for use on the RISC-V architecture (AI fog).
oPredictive analysis system for electrical equipment faults (vibration sensors).
oSystem for analyzing data from metering devices (MD) for electricity, water, heat, and other resources (metering devices).
oVideo analytics platform for compliance with technical regulations.
oSoftware package "Adaptive module based on artificial intelligence technologies for selecting the optimal technological mode of operation of an oil and gas field."
·2023 – Have applied and won a biggest Russia RnD project for AI (research direction – AIoT), with a total amount of around 1 billion rubles both from Russia government and industrial partner: https://spbu.ru/news-events/novosti/spbgu-zaymetsya-razrabotkoy-platformy-silnogo-ii-interneta-veschey-dlya (project head), first in Russia project / AI-center direction for AIoT.
·2023 - Developed Micro-RnD strategy (described below) and worked to attract external funding from industrial partners. Result: 6 Micro-RnDs with industrial partners completed.
Confidential Chinese industrial partner:
oAI-based MILP solver.
oTask Offloading in distributed OS.
oAutomatic bottleneck detection in OS.
oComputational graph forecasting for heterogeneous computing.
ООО "Gazpromneft NTC":
oReconstruction of gas curve for gas fields.
oOptimal control for gas equipment.
·2023 – Founded AI-industry competition club in SPbU (in particular AI-industry competition scholarship from SPbU). One of key results: TOP 1 in the world in one industry AI competition (team “Butsa I tudent in spbu”, I am a scientific supervisor): https://www.drivendata.org/competitions/7/pump-it-up-data-mining-the-water-table/leaderboard/.
Stack: Python, Jira (or YouTrack), GitLab, MS Project.
2015 – present: Professor of St. Petersburg State University (before 2023 was Associate Professor and Assistant Professor).
Dr.Sc. (at 29 years old) and Professor (at 30 years old) in Saint-Petersburg State University (SPbU). Created the first Russian online course on 3 languages (including Chinese language) that was broadcasted in China. H-index: 17. Number of research papers: more 90.
Responsibilities: Conducting research with students and PhD students, publishing research papers, teaching.
Achievements:
·2024 – H-index: 17. Total number of research papers: more 90. Number of Q1 papers: more 11.
·2023 – Became Professor (at 30 years old) in Saint-Petersburg State University (SPbU).
·2022 – Became Dr.Sc. (Doctoral thesis defense, second degree after PhD degree in Russia) at 29 years old in Saint-Petersburg State University (SPbU).
·2016 – Won the President of Russia scholarship for PhD students (1 of 10 winners in Russia).
2019 – 2022 – Team leader / Principal Engineer, Saint-Petersburg Research Center Huawei (RnD).
First member (started) of Autonomous Driving Network Lab, SRC Huawei, hired first team of 7 members (2 - PhD, senior RnD engineers, 5 – middle and junior RnD engineers).
Responsibilities: Managing key RnD projects of Huawei Russia in the areas of optimization for telecommunications; creating telecommunication laboratory in St. Petersburg RnD office. Managing energy direction of RnD projects, developing energy control systems, optimizers for energy direction.
Achievements: Creating telecommunication laboratory in the St. Petersburg RnD office, creating energy direction in St. Petersburg RnD office. Development of close to optimal (energy cost) energy management system.
Stack: Python, Jira, Confluence, GitLab.
2015 – 2017 – RnD Engineer, SPb Research & Development Department (RnD) in Siemens.
Received first experience in RnD, ideas generation, testing, learning how RnD is connected to startup ideas (learned ?Lean StartUp?).
Responsibilities: Working on RnD projects for automatic energy lines and equipment design systems (electricity lines checking and optimization) and energy plan optimization for EU (MILP model generation and optimization), product concept development and analytics (?lean startup? strategy application for RnD project development in company), development of basic mathematical algorithms (heuristic optimization algorithms and MILP methods), development of system prototypes.
Achievements: Development of hydropower plant model for EU energy plan optimization.
Stack: Java, Python, Gams, Jira, Confluence, Git.
2013 – 2015 – Data science engineer, OAO Bank Saint Petersburg.
Building data analytics for liquidity control department of OAO Bank Saint Petersburg.
Responsibilities: Improving statistical models for forecasting time series of crisis liquidity scenarios. Building multi-scenario forecasting models dependent on various factors. Coordination/protection of statistical forecasting models in the Treasury Directorate. Maintenance of software for calculating forecasting models for a crisis scenario. Developing functionality for quality testing of crisis scenario liquidity forecasting models. Daily comparative analysis of projected cash flows by instruments with actual data. Regular back testing of crisis scenario models. Participation in short-term liquidity management. Interest rate forecast.
Achievements: Building automatic forecasting software for key indicators and parameters for liquidity control of OAO Bank Saint Petersburg.
Stack: Python, Jira, Confluence.
Education:
Dr.Sc. (at 29 years old) in Saint-Petersburg State University (SPbU).
·2022 - D.Sc. of Physical and Mathematical Sciences in the specialty 2.3.1. System analysis, control and information processing, youngest doctor of science at Saint-Petersburg State University. (Thesis on "Dynamic and continuous updating of information in control models"). Presumably the youngest Dr.Sc. (at 29 years old) in Saint-Petersburg State University.
·2014 – 2017 Ph.D. student of the Faculty of Applied Mathematics and Control Processes, Saint-Petersburg State Universit. Result: in 2017 Ph.D. dissertation in applied mathematics, Saint-Petersburg State University.
·2009 – 2014 - Student of the Faculty of Applied Mathematics and Control Processes, Saint-Petersburg State University. Result: in 2014 honors degree as specialist in Mathematics, Saint-Petersburg State University (graduate work: "Non-antagonized command network game").
Key projects:
Fullly in charge of projects estimated at approx. 1 billion rubles from 2018 to 2024.
·Applied and won the biggest Russia government RnD project for AI (research direction – AioT), totaling at around 1 billion rubles both from government from industrial partner (https://spbu.ru/news-events/novosti/spbgu-zaymetsya-razrabotkoy-platformy-silnogo-ii-interneta-veschey-dlya), project head.
·Chinese company industry project (confidential industrial partner) Project: “Energy and latency aware scheduling of mobile applications execution sharing”, 2023, duration 12 months, project head.
·Chinese company industry project (confidential industrial partner) Project: “Optimal binary learning without model architecture constraints”, 2023, duration 12 months, project head.
·Government Research project : Artificial Intelligence and Data Science: Theory, Technology, industry and Interdisciplinary research and Applications: 2022-2024, St. Petersburg State University AI Center, project head.
·Chinese company industry project (confidential industrial partner) Project: ‘Combined cooling, heating and power supply’, 2022, duration 10 months: project head.
·Chinese company industry project (confidential industrial partner) Project: “Multi-agent reinforcement learning approaches for graph computation in the OS kernel”, 2021, duration 8 months, project head.
·Grants Council of the President of the Russian Federation (MK-4674.2021.1.1), "Development of a class of inverse optimal control problems with continuous information updating for tasks of an intelligent electronic driver assistance assistant", duration 2 years, 2021 - Head.
·Russian Science Foundation grant (No18-71-00081), "Development of methodological tools of the mathematical theory of sustainable agreements for the creation of digital intelligent contract formation systems, as a step towards artificial intelligence in this area", duration 2 years, 2018 - Head.
Key recent research papers in AI (last 3 years):
Author of more than 90 research papers, including more than 7 of Q1 papers.
·Under review - Multi Type Mean Field Reinforcement Learning for Optimal Resource Allocation in Heterogeneous Network. Engineering Applications of Artificial Intelligence (Q1).
·Under review - Jumping Knowledge Graph Attention Network for Beamforming Design in Wireless Cellular System. Computer Journal (Q2).
·Under review - Automated feature engineering based on explainable artificial intelligence for time series forecasting. Engineering Applications of Artificial Intelligence (Q1).
·Under review - DKDL-Net: A Lightweight Bearing Fault Detection Model via Decoupled Knowledge Distillation and Low-Rank Adaptation Fine-tuning, Scientific reports (Q1).
·Under review - Continuous Bayesian updating. Science China Mathematics (SCI Q1).
·Accepted - Multi-Agent Local Voting Protocol For Online DAG Scheduling, ICOMP, 2024.
·Accepted - Heterogeneous Computational Scheduling using Adaptive Neural Hyper-heuristic, Doklady Mathematics (Q2), 2024.
·2025 - Edge Feature Empowered Graph Attention Network for Sum Rate Maximization in Heterogeneous D2D Communication System. Neurocomputing (Q1).
·2024 - DAGCN: hybrid model for efficiently handling joint node and link prediction in cloud workflows. Applied Intelligence (Scopus: Q2, Impact Factor: 1.193)
·2024 - Resource allocation in heterogeneous network with node and edge enhanced graph attention network. Applied Intelligence (Scopus: Q2, Impact Factor: 1.193).
·2024 - Comparison of multi-step forecasting methods for renewable energy. Energy Systems (Scopus: Q2).
·2024 - Long-Term Forecasting of Air Pollution Particulate Matter (PM2.5) and Analysis of Influencing Factors. Sustainability (Scopus: Q1).
·2024 - Graph attention network enhanced power allocation for wireless cellular system. Informatics and Automation.
·2023 - Multi-agent Reinforcement Learning-based Adaptive Heterogeneous DAG Scheduling. ACM Transactions on Intelligent Systems and Technology (Scopus: Q1, Impact Factor: 3.602).
·2023 - Forecasting online adaptation methods for energy domain. Engineering Applications of Artificial Intelligence (WoS: Q1, Impact Factor: 6.215).
·2023 - Research on Robust Audio-Visual Speech Recognition Algorithms. Mathematics.
·2023 - Microgrid control for renewable energy sources. Vestnik of St Petersburg University. Applied Mathematics. Computer Science. Control Processes.
·2023 - Differential game model of resource extraction with continuous and dynamic updating. Journal of the Operations Research Society of China (SCI: Q1, Impact Factor: 1.4).
·2022 - Optimizing DSO Requests Management Flexibility for Home Appliances Using CBCC-RDG3. Computation.
·2022 - Applied routing problem for a fleet of delivery drones using a modified parallel genetic algorithm. Vestnik of St Petersburg University. Applied Mathematics. Computer Science. Control Processes.
Key conferences (past 3 years):
Participated in more than 45 conferences totally: in more than 15 industry conferences and more than 30 research conferences.
Industry conferences
Organized
2023 – Joint scientific and technical conference - round table of Lukoil-Engineering LLC and St. Petersburg State University "Service support of oil and gas production processes".
Participated
Russia Federation government: ST. PETERSBURG INTERNATIONAL ECONOMIC FORUM - 2023, 2024.
Sberbank: AI-Journey – 2020, 2021, 2022.
China company confidential partner:
·Lake Baikal Summit - 2023, 2024.
·Software/Hardware Code Design and Performance Optimization Summit - 2023.
·AI & Media Innovation Summit - 2023.
·Future Network Frontier - 2023.
·Russian Autonomous Network Forum - 2023.
·International conference on advances in computing technologies and artificial intelligence - 2022.
GazpromNeft: II Communication Session of Gazpromneft STC and partner universities - 2022 (https://leader?id.ru/events/351562), Novosibirsk, Russia.
Research conferences
Organized
2023 – 7-th International Scientific Conference ?Intelligent Information Technologies for Industry?: workshop ?Application of Mathematical Methods in Artificial Intelligence?.
2020 - Organized the "Workshop on Differential Games" and presented the report at the biggest annual conference on game theory, "StonyBrook conference on Game Theory", Stony Brook (New York), USA (online) (http://www.gtcenter.org/?page=Workshops.html).
Participated
·2023 - "Explainable AI: Graph Based Sampling Approach for High Dimensional AI System", 7th International Conference "Intelligent Information Technologies in Engineering and Production", St. Petersburg.
·2023 - "Audio-visual multi-modal meeting recording system", 7th International Conference "Intelligent Information Technologies in Engineering and Production", St. Petersburg.
·2023 - "Resource Allocation in Heterogeneous Network with Supervised GNNs", ?The Fourteenth International Conference on Swarm Intelligence? (ICSI'2023), Shenzhen, China.
·2023 - "Monte Carlo Tree Search with Adaptive Estimation for DAG Scheduling", ?The Fourteenth International Conference on Swarm Intelligence? (ICSI'2023), Shenzhen, China.
·2022 - “Operations Research in Industrial Problems: Adaptive Control, Optimization, Prediction, Anomaly Detection”, All-Russian conference of natural sciences and humanities, international participation in “Science of Spbu - 2022” (https://events.spbu.ru/events/science?2022), Saint-Petersburg, Russia.
·2022 - The 4th International Conference on Machine Learning and Intelligent Systems (MLIS 2022), Seoul, Republic of Korea.
·2022 - Intelligent Systems Conference 2022 (IntelliSys 2022), Amsterdam (online).
·2022 - “Optimizing Distribution System Operator(DSO) Requests Management Flexibility for Home Appliances using Contribution-Based Cooperative Coevolution Recursive Differential Grouping 3 (CBCC-RDG3)”, Mathematical Optimization Theory and Operations Research (MOTOR2022), Petrozavodsk, Karelia, Russia.
·2022 - “Comparison of Reinforcement Learning based Control Algorithms for Autonomous Driving Problem”, Mathematical Optimization Theory and Operations Research (MOTOR2022), Petrozavodsk, Karelia, Russia.
·2022 - III International Conference on Neural Networks and Neurotechnologies (NeuroNT’ 2022), Saint-Petersburg, Russia.
·2020 - IFAC2020: World Congress on Optimal Control, Berlin, Germany (online).
·2020 - "Workshop on Differential Games "StonyBrook conference on Game Theory", Stony Brook (New York), USA (online),
Teaching experience:
Launched the new international Master program, new international online course, extensive teaching experience in machine / deep learning, mathematical statistics, optimization and operations research.
·2022 – Created the first twin diploma Master Program between Saint-Petersburg State University and a university from China where students study in Russia (for 2 years): ?Mathematical and Artificial Intelligence Methods for Digital Economy?.
·2021– present - Founded the SPbU Industry AI Competitions club (students are preparing for future competitions and organized together with their professors). Link to the group: https://vk.com/club202108877
·2018 - Set up the first Russian online course “Mathematical game theory” to be broadcasted in Chinese (for Xuetangx.com platform). Generally, this online course has 3 languages (Russian, English, Chinese) and it is based on 2 platforms: Coursera.org and Xuetangx.com:
ohttps://www.coursera.org/learn/matematicheskaya?teoria–igr
ohttps://www.coursera.org/learn/mathematical?game–theory
ohttps://next.xuetangx.com/course/spbu07011001778/1075632
·2015 – present - Teaching at St.-Petersburg State University:
oMachine learning (lectures and practice),
oDeep learning (lectures and practice),
oMathematical Statistics (practice),
oOperations Research and Optimization (practice),
oDiscrete and Probabilistic Models (lectures, seminars, practice),
oLogistics and Inventory Management (seminars).
·2017 – 2019 - Teaching at National Research University Higher School of Economics:
oMathematical Economics and Statistics (lectures, practice),
oGame Theory and its Applications (lectures, practice),
oCalculus of Variations (lectures, practice).
Other:
·2017 - Head, StartUp group for development of mobile App "LifeCatch" (cross-platform solution for automation of photo processing, photo visualization and album creation). Role in the project: Head of the group of 5 developers and designers, developer of mathematical algorithm. (https://play.google.com/store/apps/details?id=com.mycompany.travellante).
Web-links with related info:
·https://spbu.ru/news-events/novosti/pmef-2024-ekspert-spbgu-po-ii-rasskazal-o-buduschem-tekhnologii-umnogo-doma
·https://spbu.ru/news-events/novosti/spbgu-zaymetsya-razrabotkoy-platformy-silnogo-ii-interneta-veschey-dlya
·https://spbu.ru/centr-iskusstvennogo-intellekta-i-nauki-o-dannykh-spbgu
·https://journal.spbu.ru/#flipbook-df_13998/21/
·https://ac.gov.ru/news/page/rossijskie-ucenye-rasskazali-o-razrabotkah-v-sfere-ii-27748
·https://gas-forum.ru/ru/novosti/iskusstvennyj-intellekt-ne-serebrjanaja-pulja-no-instrument-dlja-proektov-neftegaza/
