HSE Develops Its Own MLOps Platform
HSE researchers have developed an MLOps platform called SmartMLOps. It has been created for artificial intelligence researchers who wish to transform their invention into a fully-fledged service. In the future, the platform may host AI assistants to simplify educational processes, provide medical support, offer consultations, and solve a wide range of other tasks. Creators of AI technologies will be able to obtain a ready-to-use service within just a few hours. Utilising HSE’s supercomputer, the service can be launched in just a few clicks.
Many researchers working in AI development lack sufficient experience in building web services, managing distributed computing resources, system administration, information security, and deployment process automation. SmartMLOps was created to help eliminate the need to learn additional technologies and to address issues of repeatability and reproducibility in machine learning experiments.
The project partners include the HSE AI Research Centre, the School of Software Engineering at the HSE Faculty of Computer Science, and the HSE Tikhonov Moscow Institute of Electronics and Mathematics (MIEM). Experienced and highly qualified HSE specialists are involved in the development and deployment of the MLOps platform.
According to the project lead, Hadi Saleh, the platform is already undergoing pilot testing. Its primary goal is to reduce the time and resources required to deliver product-level services. SmartMLOps ensures the necessary level of reliability, security, and transparency in the use of AI technologies. It allows developers to save time on non-core activities such as system administration, DevOps, and CI/CD. The platform provides a unified graphical interface through which users can create a repository with an AI module template, access models, a monitoring system, orchestrators, and the supercomputer, as well as track their services’ metrics and promptly address issues.
To create a multifunctional, user-friendly, and modern service, the platform’s developers tackled the complex challenge of integrating HSE’s cHARISMa supercomputer cluster into cloud computing. The platform is based on a microservice architecture, with the main component being a framework developed by the HSE AI Research Centre.
Hadi Saleh
‘The platform's functionality includes the creation, storage, versioning, training and fine-tuning of AI models; the deployment of ready-made models as product services with REST APIs; the construction of training pipelines; the monitoring of services and the overall state of computing resources; the ability to use data from HSE’s corporate information systems; and access to cloud storage and supercomputer resources. The framework itself and the first AI modules have already been deployed, and analytical dashboards for monitoring the platform’s status have been implemented. A special autoscaling technology was developed and introduced to save resources,’ said Hadi Saleh.
Unlike commercial solutions such as Amazon SageMaker or Google AI, SmartMLOps is primarily aimed at researchers and students at HSE. To gain access, they simply need to submit an application and await approval. Until the end of 2025, anyone interested may take part in the platform's pilot use.
A special ETL process was developed to securely extract and anonymise information from HSE’s corporate systems. This helps in analysing student behaviour and building predictive models, for example, to forecast career prospects.
Elena Kozhina
‘We expect that the implementation of the SmartMLOps system will mark a significant milestone in accelerating the development of AI-based services at our university,’ noted Elena Kozhina, Deputy Vice Rector and project curator.
At present, eight AI models have been deployed on the platform, and integration has been carried out with three of HSE’s corporate information systems. Two subsystems of the MLOps platform, developed with the involvement of specialists from the Faculty of Computer Science and MIEM, have been registered with the Russian Federal Service for Intellectual Property (Rospatent).
Sergey Lebedev
‘The development of such an ambitious and promising infrastructure project posed a considerable challenge and a timely necessity for the university. SmartMLOps enables the simultaneous solution of research, educational, and administrative tasks of various scales and levels. We managed to form a unique team of professionals and involve students and research assistants in the process,’ said Sergey Lebedev, Head of the School of Software Engineering at the HSE Faculty of Computer Science.
Anton Sergeev
Anton Sergeev, Director of the Centre for Software Development and Digital Services at HSE MIEM, emphasised that systematic work with AI models is impossible without MLOps platforms, which support the full lifecycle of models—from data import and cleaning to controlling the outcomes. ‘Such high-tech projects set a benchmark for AI leadership not only within our university, but across the country as a whole. They represent HSE’s scientific excellence in the form of a system that will assist ML specialists in applying AI models to a wide range of real-world problems,’ he said.
Pavel Kostenetskiy
Pavel Kostenetskiy, Head of the Supercomputer Modelling Unit, gave the service high praise. ‘The MLOps platform will make it easier for HSE researchers working in AI to access the ‘cHARISMa’ supercomputing cluster and will lower the entry threshold into high-performance computing. In a sense, HSE’s supercomputer will appear like a cloud within this MLOps platform, which many will find convenient,’ concluded Pavel Kostenetskiy.
See also:
Centre for Language and Brain Begins Cooperation with Academy of Sciences of Sakha Republic
HSE University's Centre for Language and Brain and the Academy of Sciences of the Republic of Sakha (Yakutia) have signed a partnership agreement, opening up new opportunities for research on the region's understudied languages and bilingualism. Thanks to modern methods, such as eye tracking and neuroimaging, scientists will be able to answer questions about how bilingualism works at the brain level.
AI Predicts Behaviour of Quantum Systems
Scientists from HSE University, in collaboration with researchers from the University of Southern California, have developed an algorithm that rapidly and accurately predicts the behaviour of quantum systems, from quantum computers to solar panels. This methodology enabled the simulation of processes in the MoS₂ semiconductor and revealed that the movement of charged particles is influenced not only by the number of defects but also by their location. These defects can either slow down or accelerate charge transport, leading to effects that were previously difficult to account for with standard methods. The study has been published in Proceedings of the National Academy of Sciences (PNAS).
‘Services Must Be Flexible’: How Governments Can Use Artificial Intelligence
The HSE International Laboratory for Digital Transformation in Public Administration held a roundtable titled ‘Artificial Intelligence in Public Administration: Current Trends.’ Scholars from Israel, China, and Russia discussed which public services AI can enhance and what key factors must be considered when adopting new technologies.
Artificial Intelligence Improves Risk Prediction of Complex Diseases
Neural network models developed at the HSE AI Research Centre have significantly improved the prediction of risks for obesity, type 1 diabetes, psoriasis, and other complex diseases. A joint study with Genotek Ltd showed that deep learning algorithms outperform traditional methods, particularly in cases involving complex gene interactions (epistasis). The findings have been published in Frontiers in Medicine.
Artificial Intelligence as a Catalyst for Sustainable Development
Artificial intelligence is transforming every aspect of life, expanding both our capabilities and our boundaries. At the same time, it presents new challenges for humanity, including concerns about safety, ethics, and environmental sustainability. Today, each neural network leaves a significant carbon footprint. However, with responsible management, AI has the potential to benefit the planet and become a cornerstone of a sustainable future economy. Panos Pardalos, Academic Supervisor of the Laboratory of Algorithms and Technologies for Network Analysis at the HSE Campus in Nizhny Novgorod, emphasised this point as he addressed the XXV Yasin (April) International Academic Conference on Economic and Social Development.
‘HSE’s Industry Ties Are Invaluable’
Pan Zhengwu has spent the last seven years at HSE University—first as a student of the Bachelor’s in Software Engineering and now in the Master’s in System and Software Engineering at the Faculty of Computer Science. In addition to his busy academic schedule, he works as a mobile software engineer at Yandex and is an avid urban photographer. In his interview with the HSE News Service, Zhengwu talks about the challenges he faced when he first moved to Russia, shares his thoughts on ‘collaborating’ with AI, and reveals one of his top spots for taking photos in Moscow.
Scientists Present New Solution to Imbalanced Learning Problem
Specialists at the HSE Faculty of Computer Science and Sber AI Lab have developed a geometric oversampling technique known as Simplicial SMOTE. Tests on various datasets have shown that it significantly improves classification performance. This technique is particularly valuable in scenarios where rare cases are crucial, such as fraud detection or the diagnosis of rare diseases. The study's results are available on ArXiv.org, an open-access archive, and will be presented at the International Conference on Knowledge Discovery and Data Mining (KDD) in summer 2025 in Toronto, Canada.
Megascience, AI, and Supercomputers: HSE Expands Cooperation with JINR
Experts in computer technology from HSE University and the Joint Institute for Nuclear Research (JINR) discussed collaboration and joint projects at a meeting held at the Meshcheryakov Laboratory of Information Technologies (MLIT). HSE University was represented by Lev Shchur, Head of the Laboratory for Computational Physics at the HSE Tikhonov Moscow Institute of Electronics and Mathematics (HSE MIEM), as well as Denis Derkach and Fedor Ratnikov from the Laboratory of Methods for Big Data Analysis at the HSE Faculty of Computer Science.
AI vs AI: Scientists Develop Neural Networks to Detect Generated Text Insertions
A research team, including Alexander Shirnin from HSE University, has developed two models designed to detect AI-generated insertions in scientific texts. The AIpom system integrates two types of models: a decoder and an encoder. The Papilusion system is designed to detect modifications through synonyms and summarisation by neural networks, using one type of models: encoders. In the future, these models will assist in verifying the originality and credibility of scientific publications. Articles describing the Papilusion and AIpom systems have been published in the ACL Anthology Digital Archive.
HSE Researchers Develop Python Library for Analysing Eye Movements
A research team at HSE University has developed EyeFeatures, a Python library for analysing and modelling eye movement data. This tool is designed to simplify the work of scientists and developers by enabling them to efficiently process complex data and create predictive models.