DBaaS, DRaaS, AIaaS and other promising segments of the Russian cloud market

In 2025, the division of cloud services into IaaS, PaaS, SaaS is still relevant, such a basic model is convenient for describing responsibility and level of abstraction. This gives customers a clear distinction between areas of responsibility.

As other, most promising (hybrid) segments, experts name, first of all, DBaaS (database as a service, "database as a service"), DRaaS (disaster recovery as a service, "disaster recovery or disaster tolerance as a service") and AIaaS (artificial intelligence as a service, "artificial intelligence as a service").

DRaaS helps return the system to its target in a predetermined (very short) time frame. There are different strategies here: a backup and auto-recovery service, mirrors with a full copy of the application, and much more.

The prospects of the direction are due, among other things, to growing cyber threats. After 2022-2023, interest in backup, failover scenarios, and the rapid reorientation of business services increased. DRaaS allows the business to comply with the requirements of regulators and avoid losses in case of failures without building its own backup sites. The benefits were quickly assessed in the financial sector (uninterrupted operation of payment systems), power (protection of critical infrastructure) and the public sector (continuity of public services).

According to the Telecom Exchange report, according to a survey of 125 representatives of IT areas of companies from the field of finance, telecommunications, online trading, etc., in 2024, 26% of Russian companies did not use data backup services. In this group, the average duration of downtime in the event of cyber incidents and other force majeure situations is 10 hours, in other companies - 4 hours.

Another 39% of Russian companies do not copy more than a quarter (25%) of the total amount of data. Backups from 50% to 75% of the data are made by 23% of the companies surveyed. 6% of respondents said that 100% of data is continuously copied in their companies.

According to Dmitry Denshchikov, director of cloud services at the Telecom Exchange IT integrator, there is a known case when a delivery service was attacked by a ransomware virus, as a result of which the company lost access to its own databases. As a result - a simple site and application with a length of several days and losses of several million rubles.

However, there is a positive trend here - after several high-profile incidents last year, the request for DR-loop services (Disaster Recovery), that is, for storing archives and backups at sites other than the main ones, where the customer's main infrastructure is located, has approximately doubled, "the expert said.

IAaaS is an infrastructure and services for artificial intelligence tasks. Experts talk about sustainable demand from banks, retail, logistics and industry: increasingly, businesses need cloud capacity with GPUs, the ability to train and use models, ready-made frameworks. Customers want to not just "train the model," but integrate it into the business process - and they are doing it more and more often in the cloud without deploying expensive infrastructure. Some of the most dynamically developing areas are AIaaS for medicine (medical image analysis and diagnostics), retail (demand forecasting) and industry (production optimization).

DBaaS is in demand, including in logistics (real-time supply chain management), telecom (processing of big subscriber data) and gaming (scalability of multi-user platforms).

As for the nuances of segmentation, as Denis Afanasyev, head of the cloud direction of CorpSoft24, said, most modern services are hybrid. That is, DBaaS formally refers to PaaS, but according to the payment and operation model, it can be closer to SaaS. In practice, many providers combine levels in a single sentence. Therefore, IaaS/PaaS/SaaS remains a useful scheme, but for niche services it is critical to understand at what level the client expects manageability and value.

AI as a Service (AIaaS)

AIaaS is not a single service, but a multi-level structure, each of the layers of which is at a different stage of maturity in Russia. In general, the segment cannot yet be called mature, but its growth rate is one of the highest in the cloud market of the Russian Federation (especially in 2025). Intelligent automation of business processes, recommendation systems and personalization, virtual assistants, support for frameworks, HR recruiting, QA models to improve client experience, chat bots and virtual assistants, neural network training and cybersecurity - the segment can potentially and is already developing a very wide range of cloud services and solutions.

Business in Russia is actively interested in the capabilities of AI to optimize processes, increase efficiency and create new products and services. The market is looking for P & L-confirmed benefit from AI adoption. As soon as it is obvious, the segment will experience explosive growth. There is already a high growth rate in demand for AI cloud services - about 30-35% per year. AIaaS offers become more interesting in a situation of a large shortage of AI professionals.

The supply structure in the AIaaS market is as follows: there is an infrastructure level (GPU as a service, TPU, high-bandwidth storage), platform (ML Platform, ML Ops). At the application level, there are API services (Computer Vision, NLP, Text-to-Speech - ready-made AI models) and domain solutions (industry AI). Demand for the latter is formed by retail, the financial industry, industry. Market leaders are SberCloud with Kandinsky and Gigachat products, and Yandex Cloud with YandexGPT. The leaders offer a wide range of pre-trained models and tools for machine learning. The rest of the market participants are niche players who focus on developing specialized AI solutions for specific industries or tasks.

The lease of computing power with graphics accelerators is called the most mature and large segment of AIaaS in Russia. Its volume, according to VK Cloud, in 2024 amounted to ₽17,1 billion. This is the foundation on which all other AI solutions are built. The platform level is a set of tools for data scientists to automate the entire cycle of working with models: from data preparation to training, deployment and monitoring (ML training and model deployment). This increases the added value and facilitates development, which is relevant not only due to the evolutionary growth of interest in its own LLM and CV models, but also due to the lack of corporate access to Western services: Google Vertex AI, AWS SageMaker, Azure ML.

This segment is actively developing within the framework of PaaS offers of leading providers (development platform). At the application level - pre-trained models available by API for solving specific problems (speech recognition and synthesis, text analysis, predictive analytics and forecasting). This level is closest for companies without their own ML teams. The requests of large corporations are more focused on the infrastructure level. Medium-sized companies are more often interested in ready-made modules, small commands are simple APIs for integration.

Denis Afanasyev, head of the cloud direction of CorpSoft24, noted that this niche has significant barriers: the high cost of GPU infrastructure is not affordable for every industry player. It also requires its own expertise in managed pipelines, MLOps. The demand for these types of services is also quite narrow, the solvent audience is mainly large companies and government agencies.

Шаблон:Quote 'author=Denis Afanasyev, Head of Cloud at CorpSoft24

In the coming years, drivers will be: increasing the availability and flexibility of AI cloud services compared to On-Anticipate solutions; growing demand for business process automation and data-driven decision-making. For organizations with specialized IT specialists, two main scenarios for using AIaaS are assumed: the use of cognitive API calculations and ML frameworks to create internal services for your own needs; creation of specialized services based on AIaaS for offering on the market on a commercial basis.

The global AIaaS market, where this area is in great demand, is estimated at $20 billion (by the end of 2025) with the prospect of growth to $90-116.7 billion by 2030. In Russia, estimates vary in the range of ₽10 -20 billion.

The basis of growth is formed by rapidly developing directions, including ML Ops, generative AI (GenAI), AI Search and Document AI. These segments are expected to deliver sustained growth in AIaaS through at least 2029, maintaining strong demand dynamics and expanding the supply structure in the market. The directions of NLP, CV, process optimization can be promising.

One possible development scenario: the current boom in GPUaaS investment will lead to the emergence of powerful new domestic fundamental models (LLM). This, in turn, will become the basis for the rapid growth of platform and applied AI services in the coming years. A decisive competitive advantage will be given to the provider that controls the entire stack - from access to hardware to ready-made APIs.

Significant acceleration of the introduction of AIaaS technologies in Russian companies is expected. As the corresponding supply is formed and promoted, a significant increase in demand is possible. In 2025, the dynamics may be 50%. In the next three years, the market may grow 2-4 times. The positive scenario for the Russian Federation is limited to access to advanced chips and global models.

According to the Zendesk Customer Experience Trends Report 2025, which is based on a survey of 5,000 consumers and 5.5 thousand business representatives in 22 countries, 90% of respondents believe that AI will be able to solve 80% of problems without human participation over the next few years.^[1] On the other hand, in 5-10 years, the "era of AI fatigue" may come.

Data Recovery as a Service (DRaaS)

DRaaS (Disaster Recovery as a Service) is one of the most efficient ways to back up data that is built into your organization's IT infrastructure. DRaaS allows a business to confidently cope with any emergency, ensuring stability and stability in an unpredictable environment. Demand for DRaaS is a direct indicator of business maturity.

Many Russian providers today claim the DRaaS service as a service, but in most cases we are talking about a backup. For large and medium-sized businesses, the requirements for DRaaS have reached a new level, because companies use combinations of local infrastructure, public and private clouds. Modern DRaaS is a complex scenario of rapid recovery, testing, flexibility.

Often cloud providers provide a DR-loop service, which is only part of DRaaS. The backup service does not provide a measure of the recovery time of application systems, while DRaaS provides guarantees: for how long the final application system will work again, and what will be the permissible amount of lost data.

Thanks to DRaaS solutions, companies reduce infrastructure costs by up to 70%, as well as minimize data loss risks, financial and reputational downtime losses (30-50% reduction in downtime costs). The speed of recovery has become a competitive advantage: disaster resilience protects the trust of customers who expect their suppliers to run smoothly. The fault tolerance of businesses increases, the risks of information loss due to the actions of cybercriminals are reduced. DR management outsourcing reduces Opex: companies do not need to maintain staff, and routine management and recovery tasks are automated.

State influence

Tightening regulatory requirements in Russia has an extremely high impact on the development and implementation of DRaaS. Responsibility for data breaches is being tightened. Services shall comply with standards such as GOST R 57580. The requirements of Federal Law No. 187-FZ "On the Security of the Critical Information Infrastructure of the Russian Federation" and No. 152-FZ "On Personal Data" directly prescribe operators to ensure continuous operation of systems and rapid recovery. There are also recommendations and requirements of FSTEC and Roskomnadzor. The frequency of testing, documentation, proof of performance are important. The use of foreign solutions is limited. Providers must implement state-of-the-art encryption, access control, and threat monitoring technologies. DRaaS shall provide storage of backup copies of data in Russian data centers.

For many companies, especially in the financial sector, public administration and industry, the fuel and energy complex, the introduction of a full-fledged disaster recovery plan is mandatory. Telecom operators and participants in infrastructure industries are required to document recovery scenarios and conduct tests. These industries have become the top consumers of DRaaS.

Regulatory restrictions for the banking and insurance sectors became known back in 2023. Companies from these sectors could not use DRaaS in public clouds, the best option for them was to build private clouds or colocation.

Thus, the demand for DRaaS shifts from the plane of technical need to the plane of strategic risk management and business continuity.

DRaaS preferred characteristics

DRaaS is becoming part of a proactive business strategy that must be deeply integrated into core IT processes, support automation, and adapt to complex architectures.

Modern DRaaS must comply with regulatory requirements (FSTEKRossia, Central Bank), encrypt data during transmission and storage, have geographical distribution with data centers in different regions to minimize risks, ensure minimum target downtime (Recovery Time Objective, RTO), minimize data loss (Recovery Point Objective, RPO).

RTO should be less than an hour (minimum), ideally counting for minutes or even seconds, including through the use of technologies such as real-time data replication and snapshots.

Switching to redundant resources in the event of a failure should be fast, and the processes of switching (failover) and returning to the main site (failback) should be as automated and orchestrated as possible, with minimizing the human factor.

It is critical to conduct regular, non-destructive disaster recovery drills in an isolated environment. This is the only way to ensure that DR scripts are ready.

DRaaS solutions must be transparent, with constant access to monitoring the health of systems and data, key metrics for timely detection of threats, fault detection, analytics, and the launch of recovery procedures.

The service should be flexible: support heterogeneous and hybrid environments, including physical servers, virtual machines in the private cloud, and resources in several public clouds.

It is important to easily integrate with the existing IT infrastructure of the company, with the provider cloud for quick deployment of a backup environment, the ability to quickly scale resources.

Providers must have at least two data centers in order to place customer services on one site, and a backup copy on the second, from which the system can be restored in full. The provider needs to dive deep business processes into the customer and their systems to ensure their fault tolerance with the necessary. SLA An additional plus will be the ability to configure recovery options for specific customer needs.

DRaaS process trends

Experts note the introduction of AI/ML algorithms that allow you to proactively identify failures, minimize their consequences even before serious incidents occur (predictive analytics), and optimize automatic recovery plans. Recovery templates for typical services are considered in demand.

Given the growth of cyber threats, transparent data encryption on the fly, strict access control and the ability to isolate DR environments are increasingly required. CDP (Continuous Data Protection) is relevant - backup after each change. Zero Trust principles are beginning to be actively implemented directly into the architecture.

Integration with cloud storage (object, block), support for multi-clouds, clustering at the storage level is relevant. Containerization (DRaaS based on Kubernetes) is used to quickly deploy applications in a backup environment.

Engineers have to retrain when switching to domestic backup and replication solutions. Against the background of a fairly wide choice of Russian virtualization, there is a growing demand for such backup and replication systems that can work with different platforms. And while compatibility is not configured for everyone, there is a demand for a separate expertise that allows you to connect backup sites with different virtualization systems to each other.

Database as a Service (DBaaS)

Database as a Service (DBaaS) is one of the fastest growing and most critical components of PaaS. This is not a separate service, but rather a targeted offer for closing scenarios, which also includes MKaaS, BI and Cloud Registry.

DBaaS ceases to be just a "wrapper" around classic DBMS, today they are full-fledged platforms in which the database becomes a central element capable of performing analytics, working with vectors, training models and adapting to load in automatic mode.

The average estimate of the volume of this subsegment in the Russian Federation, according to a survey conducted by us, is ₽8 -10 billion in 2024 (with 2.7 billion in 2020). According to some forecasts, in 2025 it is expected to grow by 35-40% or to a level of over ₽12 -15 billion. The bar of ₽20 billion can be overcome in a few years. Until 2028, the CAGR may be at 16%. In another scenario, from 2024 to 2029. the sub-segment will demonstrate CAGR in the range of 24-32%. It is possible to grow to ₽25-35 billion, which will make DBaaS one of the main "locomotives" of the cloud market in Russia.

Among the technological trends, the most often mentioned are integrations with AI/ML services, Multi-Model Database (support for several models in one DBMS), serverless (auto-scaling, payment by use), hybrid solutions (allow you to combine the advantages of cloud and local databases), specialized databases (for specific business tasks), needs for analytical functionality. Almost no one bypasses the relevance of MySQL.

The development of cloud databases in Russia will depend on many factors, including the economic situation, regulatory requirements, technological innovations and the level of confidence in local solutions. Companies will gradually implement DBaaS, but maintain hybrid approaches, the share of imported solutions will decrease. The key factors may be economic instability and limited budgets of companies, a lack of qualified personnel (DBA), the gradual development of the infrastructure of Russian providers.

DBaaS attracts the attention of companies in the spectrum from SMB to Enterprise. The drivers here are classic: digitalization policy, import substitution, data growth, the need for scalable and flexible solutions, management complexity, personnel shortage (DBA), TCO reduction. As the digital economy evolves and big data accumulates, companies are actively moving to cloud-based solutions to simplify management, scale, and reduce capital expenditures.

Other promising directions in 2025

According to VK Cloud, the Russian cloud security market (Security as a Service, SECaaS) has reached ₽5 billion rubles and continues to grow. Specialized services such as managed web application firewalls (WAFs), incident monitoring and response centers (SOCs), secure development platforms (DevSecOps), DDoS protection, vulnerability scanning, data breach protection will all be in great demand. SECaaS allows companies of all sizes to access expertise and technology that was previously only available to large organizations. Separately, we can mention such areas as cloud security for banking (transaction protection), e-commerce (fraud prevention) and industrial systems (cyber protection of IoT devices).

Everything related to compliance is relevant. Compliance as a Service is progressing, enabling companies to effectively integrate compliance practices into their processes.

Cloud solutions developed taking into account the specifics of various businesses will be relevant. For example, in health care, finance, industry. So, for telemedicine, medical data management (taking into account 152-FZ) and AI diagnostics, Healthcare Cloud is relevant. For the financial sector (in accordance with 115-FZ, 152-FZ,) FSTEC Russia , highly secure Fintech Cloud platforms will be useful. Specialized certified GovCloud solutions are suitable for the public sector. Also, Blockchain as a Service will be relevant for the public sector and fintech.

Against the background of state support for the digitalization of industry, a huge demand is being formed for platforms combining the industrial Internet of Things (IIoT), predictive analytics, digital twins and production management systems. This is a capital-intensive but strategically important market.

As for computing, experts note such promising niches: FaaS (Function as a Service) for microservices; Container as a Service - merger of IaaS and PaaS, allowing developers to run container applications without infrastructure management; Edge computing - computing and data processing occur closer to the source of their occurrence (the approach minimizes latency, which is important for applications that require instant response, for example, in IoT or in streaming services). The demand for GPUaaS is still point (AI developments, scientific projects).

Clouds under "1C": according to a study by K2 Cloud, 32% of Russian organizations transfer "1C" to the cloud due to the lack of performance of their own On-Anticipate infrastructure, while it is very important to fully support the provider, guaranteed by SLA for technical support.

FinOps and GreenOps as a service: as cloud accounts grow and infrastructure becomes more complex, Russian companies will inevitably have a demand for tools and consulting services for managing cloud finance (FinOps) and environmental footprint (GreenOps).

Media & Entertainment Cloud will be relevant for content processing and delivery.

Notes