GigaCloud's Server Bill Is a Recovery Promise in Ukraine

Summary

• GigaCloud's strongest public case is not that every Ukrainian workload should stay inside Ukraine. It is that a server account can be bought as a recoverability bundle: VMware or OpenStack capacity, local support, backup and disaster-recovery services, Ukrainian and EU data-center choices, public-sector compliance claims, and tariff components priced in hryvnia.
• The evidence supports a serious domestic-cloud proposition for Ukrainian businesses and public users that care about recovery time, support language, procurement rules and data location. It does not prove that GigaCloud is always cheaper or operationally safer than AWS, Azure, Google Cloud, Oracle Cloud or other Ukrainian providers, because the public record does not disclose customer-level uptime, restore-test success rates, churn, gross margin, incident history or workload-by-workload total cost.
• The unit to analyze is the Ukrainian cloud server account: a recurring account that combines compute, memory, storage, IP connectivity, backup quota, support hours, licenses and recovery options. The tariff schedule shows why the account is a risk contract, not a single virtual machine price.
• GigaCloud's public record is unusually rich for a private regional cloud provider: Ukrainian company registry information, procurement sales signals, RIPE membership, a published tariff schedule, product pages, certification and government-cloud claims, customer case studies and technical DNS/BGP surface. Those sources point in the same direction, but they do not remove the need for buyer diligence on restore tests, power resilience, network paths and contract liability.

The server decision starts with failure, not with a cloud logo

Imagine a Ukrainian retailer, logistics company, municipality or software vendor asking a narrow question after another year of war, outage risk and budget pressure: if the production server fails tonight, where should the recoverable copy live? The cheapest answer on a spreadsheet may be a foreign hyperscaler region. The most familiar answer may be a server already sitting under an internal administrator's care. GigaCloud's answer is different. It asks the buyer to treat a cloud server account as a standing recovery promise: capacity that can be scaled, backed up, supported locally, contracted under Ukrainian commercial routines, and placed in Kyiv, Lviv or Warsaw according to the customer's risk model.

That is a more demanding claim than "local cloud is good." It has to survive three questions. First, what exactly does the customer buy beyond the label "server"? Second, why does the unit become expensive once labor, capital, compliance, outage time, exchange rates, supplier dependency and restore failure are included? Third, does the public evidence prove that the unit is worth paying for, or does it only show that GigaCloud has built a credible sales argument?

The unit in this article is the Ukrainian cloud server account. It is not a bare virtual machine in isolation. In GigaCloud's public materials, the customer can buy public cloud on VMware, public cloud on OpenStack, dedicated or private cloud, GPU cloud, backup, replication, Veeam services, support hours, Microsoft and Cisco licenses, public IP addresses, internet connectivity, cross-site channels, and services for regulated or public-sector workloads. A recoverable account is therefore the bill that appears when a customer adds enough of those items to make a server useful in a crisis: processor and memory, storage class, backup quota, replication service, internet path, a support arrangement, and sometimes a compliance wrapper.

That is why the price cannot be judged only by comparing the monthly cost of vCPU and RAM with a foreign on-demand instance. A Ukrainian business that needs a recoverable server is paying to avoid a larger loss: data that cannot be restored, a tax or payment service that cannot open, a retail stock system frozen during a sale period, a municipal service that cannot retrieve records, or a software platform whose support team cannot reach the provider in the language, time zone or contract form it needs. GigaCloud's thesis is valuable only if its local and regional architecture converts those risk costs into lower expected failure cost.

The available evidence supports the existence of that proposition. GigaCloud publishes a tariff schedule effective from October 1, 2025. Its official product pages describe five technical sites in Ukraine and the EU, including Kyiv, Lviv and Warsaw. It claims 99.95 percent SLA for VMware-based and dedicated cloud services and 99.5 percent for its OpenStack public cloud. It publishes backup and disaster-recovery pages that discuss RTO and RPO, Veeam-based backups, protected repositories, and cloud storage in Ukraine or the EU. It says its infrastructure has security certifications and government-cloud approvals, including references to Ukraine's State Service of Special Communications and Information Protection. Opendatabot records the company as a Ukrainian limited liability company with code 39792589, Kyiv address, 2025 revenue of 561.443 million UAH and 2025 net profit of 130.259 million UAH. It also shows large public-procurement sales to public customers, with DP Diia as the largest named buyer by aggregate amount.

That evidence does not settle every business question. There is no public customer-level service ledger showing actual outages, recovery tests, failed backups, credits paid under SLA, or comparative total cost against AWS, Azure, Google Cloud, Oracle Cloud, De Novo, Parkovy, Tet, hosted private infrastructure or in-house equipment. The public case is therefore not "proven superiority." It is narrower and more interesting: GigaCloud has enough public evidence to be judged as a serious local substitution option, but the buyer should value the account only where recovery, data location, support and procurement fit matter more than the cheapest undifferentiated compute.

GigaCloud is a Ukrainian cloud company, not just a RIPE entry

The company analyzed here is Gigacloud LLC, and the public identity behind it is GigaCloud, a Ukrainian cloud provider. Opendatabot lists the full Ukrainian legal name as the limited liability company "GIGACLOUD", with English name "GIGACLOUD" Limited Liability Company, code 39792589, address at Vasylkivska Street 37, letter B, Kyiv, and registration date May 15, 2015. Its main activity is data processing, hosting and related activities. Opendatabot also lists additional activities including wired telecommunications, computer programming, IT consulting, web portals and other IT services.

The company also appears in RIPE NCC's public member list for Ukraine as Gigacloud LLC, with the same Kyiv address. That matters for public surface, because it shows the company is not merely reselling a web brand without number-resource governance context. It does not prove network quality or data residency, but it does show that GigaCloud has an identifiable footprint in the internet infrastructure administrative layer.

Ownership and group context are more complex than the product pages alone suggest. Opendatabot lists founders including Nazlami Limited in Malta, Telliani Limited in Cyprus and Zvityaga Limited in Malta, and lists Ukrainian ultimate beneficial owners including Oleg Polishchuk and Nazariy Kurochko. Diia.City United describes Nazariy Kurochko as founder of GIGAGROUP, which includes national telecom operator GigaTrans, data center GigaCenter, cloud operator GigaCloud and cybersecurity agency GigaSafe. That group context is commercially important. A cloud provider that is near telecom, data-center and cybersecurity sister businesses may sell a more integrated infrastructure proposition than a pure virtual-server reseller. It also creates questions a buyer should examine: related-party dependence, data-center ownership versus leasing, network-provider concentration, and whether support responsibilities sit clearly in the GigaCloud contract.

The public financial indicators are stronger than expected for a private cloud-services company in a regional market. Opendatabot reports 2025 revenue of 561.443 million UAH, 2025 net profit of 130.259 million UAH, 2025 assets of 203.449 million UAH, and 141 employees for 2025. The same page shows revenue increasing from 109.354 million UAH in 2020 to 148.090 million UAH in 2021, 239.306 million UAH in 2022, 321.045 million UAH in 2023, 411.655 million UAH in 2024 and 561.443 million UAH in 2025. It also reports first-quarter 2026 revenue of 170.006 million UAH and 21.288 million UAH in net profit.

Those numbers should not be mistaken for audited public-company reporting. Opendatabot is an aggregator of Ukrainian company and registry data, and the article does not independently audit the underlying filings. Still, the scale is meaningful. A cloud provider with hundreds of millions of hryvnia in annual revenue and named public-sector customers is different from a small hosting shop. It has enough public scale to make reliability, compliance and procurement claims commercially testable.

The account is assembled from many small meter lines

GigaCloud's tariff schedule is the most useful economic document in the public record because it turns the server account into visible components. The October 2025 tariff PDF is an official GigaCloud document for services and tariffs, signed and stamped, with Kyiv address and company code. It introduces tariffs effective October 1, 2025 and lists service components in hryvnia, before VAT, VAT and total monthly cost with VAT.

For the OpenStack public cloud group, the visible tariff lines include Linux memory at 138 UAH per GB per month including VAT, Windows memory at 228 UAH, SSD storage at 4.80 UAH per GB, HDD storage at 1.44 UAH per GB, backup storage quota at 1.20 UAH per GB, public SFTP storage at 0.96 UAH per GB, a 100/100 Mbit/s internet connection for Ukraine and worldwide traffic at 516 UAH, a Layer 2 100 Mbit/s channel between GigaCenter and the Lviv data center at 840 UAH, and an IP address at 204 UAH. The same table includes GigaCloud Admin service hours from 1,200 UAH per hour for up to 10 hours to 1,056 UAH per hour from 50 hours, including VAT.

For VMware public cloud, the schedule shows vCPU at 216 UAH per GHz per month including VAT, vMEM for Linux virtual data centers at 288 UAH per GB, and vMEM for Windows at 345.60 UAH. It lists storage classes, including HDD up to 200 IOPS at 1.56 UAH per GB, SSD up to 3,000 IOPS at 6.00 UAH, fast SSD up to 20,000 IOPS at 8.40 UAH, and cold HDD virtual disk tiers that decline with volume. It also lists internet and Layer 2 channel components, an additional public IP address at 204 UAH, and a separate GigaCloud Admin hour table. Backup as a Service lines show Veeam Backup and Replication Enterprise Plus at 720 UAH per virtual machine, local backup space at 0.96 UAH per GB, remote backup space at 1.20 UAH per GB, Cloud Connect Backup Server at 480 UAH per virtual machine, Cloud Connect Backup VM at 360 UAH per virtual machine, and Microsoft 365 backup at 96 UAH per user plus backup storage.

This tariff structure matters more than any single price. A recoverable server account is a pile of monthly components, not one SKU. A company may start by pricing four vCPU, sixteen GB of memory and fast SSD storage, but the bill becomes economically meaningful only when the customer adds backup space, public IP, connectivity, support hours, possible Veeam licensing, cross-site channels and any compliance-specific services. The tariff also makes a currency point. GigaCloud sells Ukrainian accounts to Ukrainian legal entities in UAH. A business that earns in hryvnia and pays staff, taxes and local suppliers in hryvnia may prefer a cloud bill that does not move every month with USD or EUR exchange rates.

The tariff document does not prove that GigaCloud is cheap. It proves that buyers can decompose the bill into understandable resource and service lines. That is important in a market where hyperscaler pricing often looks cheap at the VM level and then becomes harder to forecast once egress, storage operations, support tiers, marketplace licenses, currency movement and foreign tax treatment are included. GigaCloud's own hyperscaler migration page makes this argument directly, saying foreign providers currently offer preferential terms to Ukrainian companies that could be cancelled, and presenting indicative comparisons where GigaCloud is the baseline "n" while AWS, Azure and GCP are higher for specified scenarios. That is a marketing claim and should be treated as such. But the underlying accounting issue is real: a server account is not a server; it is a monthly recovery and operations commitment under exchange-rate, labor and contract constraints.

Locality is valuable only when it changes the recovery path

Data location is easy to oversell. A server in Ukraine is not automatically safer than a server in Frankfurt, Warsaw or Dublin. During war, some Ukrainian workloads may need foreign recovery sites precisely because local infrastructure faces physical and power risks. GigaCloud's stronger argument is not pure domesticity. It is optional locality: the ability to place workloads and backups across Ukrainian and EU sites according to a customer's legal and operational need.

The official VMware public cloud page says GigaCloud hosts equipment in five data centers located in Kyiv, Lviv and Warsaw, conforming to Tier III and Tier IV reliability standards. Its FAQ says Ukrainian data can be hosted at EU sites, including Atman in Warsaw and Equinix WA3, with Equinix WA3 available when paid in euros. The "Why GigaCloud" page says equipment is located in five data centers: three in Ukraine and two in Poland. The backup and replication page says solutions are based on five technical sites in Ukraine and the EU, with Tier III and Tier IV data-center reliability. The disaster-recovery page frames BaaS and DRaaS as backup and replication in cloud storage within the EU or Ukraine.

That geography is central to the recovery promise. A Kyiv-only cloud account would be difficult to sell as the safest answer to a wartime recovery problem. A Ukraine-and-Warsaw account can be configured around different risks: low-latency Ukrainian operations, Lviv or Warsaw backup site, EU-hosted copy for foreign-market or risk-spreading needs, and a contract with a Ukrainian support team. The economic question becomes whether the customer pays for actual separation or merely for a provider that advertises multiple sites. A one-site virtual machine does not become recoverable because the provider has other sites in the brochure. The customer has to buy and test backup or replication across the sites that matter.

GigaCloud's Lviv expansion during the early phase of the full-scale invasion gives the geography more credibility. Lviv IT Cluster reported in 2022 that GigaCloud expanded its platform in the Lviv data center, set up a cloud cluster and backup storage system, and provided more than 300 TB of memory for customer backups. The article quoted a company representative saying GigaCloud had planned for force majeure and wanted Ukrainian businesses to move services and data to a safer place in case of invasion. It also described independent channels for the Lviv data center so traffic would not depend on Kyiv. This is not a neutral audit, but it is a useful contemporaneous account of why western Ukraine mattered as a resilience location.

The public record also supports a public-sector locality angle. GigaCloud's NIST-authorized cloud page says its cloud infrastructure passed security authorization under Ukraine's State Service of Special Communications and Information Protection and is listed in the official register of authorized security systems and the list of cloud service and/or data center service providers. The same page says the authorization model is based on NIST SP 800-53 and a risk-oriented approach. Its sovereign cloud page says sovereign cloud is infrastructure compliant with state regulations on data storage, processing and transmission within national borders, and says GigaCloud is included in the official SSSCIP list of cloud service providers.

The value of those claims depends on the buyer. A private e-commerce company may care less about official lists than about latency, restore tests and support. A state enterprise, regulated health system, financial monitoring entity or critical-infrastructure operator may care deeply about whether a provider can be used under Ukrainian procurement and security rules. Asters' February 2025 note on new Ukrainian cloud and data-center regulation explains that public users and critical-infrastructure operators face a framework for providers, electronic catalogues, model contracts, conformity documents, incident notification and, after the transition period, eligibility tied to the official list. That makes local compliance part of the account value, not just a sales badge.

The support hour is where cheap infrastructure becomes expensive

The most revealing line in the tariff schedule may be the admin hour. GigaCloud prices GigaCloud Admin service hours at 1,200 UAH including VAT for up to 10 hours and declining to 1,056 UAH from 50 hours. This looks like a small services line next to cloud capacity, but it exposes the hidden cost of recovery.

Servers fail in ways that require labor. Someone has to design the backup policy, confirm that the correct volumes are included, manage keys, configure network access, test restoration, patch the system, adjust memory and storage after usage changes, and coordinate with application owners during a migration. If a business buys a cheap foreign VM and then spends internal senior engineering time learning billing, networking, IAM, backup tooling and support escalation, the VM is not cheap. It has merely shifted cost from the provider invoice to the payroll, failure queue and after-hours incident process.

GigaCloud's official pages lean heavily into this labor proposition. The company says technical support accepts requests 24/7 by phone, email and customer portal. Its hyperscaler migration page emphasizes Ukrainian-language technical support and says GigaCloud provides an individual approach to each client. Its VMware page says specialists perform client service migration to GigaCloud infrastructure for free. Its backup and replication page says GigaCloud selects optimal solutions with regard to RTO and RPO. These are sales claims, but they align with a rational buyer problem: Ukrainian firms often have scarce infrastructure engineers, and a recoverable server is worth more if the provider reduces the time those engineers spend on undifferentiated infrastructure work.

The support proposition is also where evidence is thinnest. Public pages tell us support exists, not how quickly it solves incidents, how often first-line support escalates, what service credits are paid, or how much emergency work costs under specific contracts. The tariff schedule tells us admin hours are metered and not free. That matters. A buyer should not hear "local support" as unlimited engineering bundled into every VM. The economic account should model likely support hours, migration work, backup design and occasional restoration assistance.

Still, local support can be a genuine economic advantage even when it costs money. Time zones, language, contract law and procurement routines matter during a crisis. If a public agency or mid-sized Ukrainian company cannot navigate a global support queue quickly, a foreign provider's lower compute line may be offset by incident labor and delay. The public evidence supports GigaCloud's claim that it sells not only capacity but also managed infrastructure services. It does not prove the service quality level, so the account is worth paying for only if the buyer tests the support process before the incident.

Backup discipline is the core product, not an add-on

The business thesis turns on backup discipline, and GigaCloud's public pages provide enough material to test it. The backup and replication page describes BaaS and DRaaS as two different products. BaaS creates and stores backup copies of IT infrastructure in cloud storage and can cover infrastructure hosted in third-party clouds, physical servers and individual computers. DRaaS is described as a fully functional backup virtual data center, built using Veeam or VMware vCloud Availability, to which the client's IT infrastructure is replicated together with deployed applications. The page says operations can be switched to the backup environment within minutes after failure or primary-infrastructure unavailability, and that DRaaS is distinguished from standard backup by faster RTO and lower RPO.

The tariff schedule makes this more concrete. Backup costs appear as storage, software license and service components. Veeam Backup and Replication Enterprise Plus is listed at 720 UAH per virtual machine for one BaaS configuration. Local backup space is 0.96 UAH per GB; remote backup space is 1.20 UAH per GB. Veeam Cloud Connect Backup VM is listed at 360 UAH per virtual machine. Microsoft 365 backup is 96 UAH per user plus backup space. These numbers show why the server account becomes expensive. A buyer who wants meaningful recovery cannot price only compute and memory. It must pay for backup software, storage volume, remote location, configuration labor and tests.

Company case studies show how GigaCloud wants buyers to understand this. Its Delta Sport case says the retailer uses a multi-layer backup strategy aligned with the 3-2-1 rule, with backups stored locally, in the GigaCloud cloud and off-site. The Prozorro case says the public procurement platform had moved from AWS to a Ukrainian data center due to Ukrainian legislation, then decided to operate across two data centers to improve fault tolerance. GigaCloud says it built an isolated private cloud on OpenStack for Prozorro, distributed components between data centers, configured monitoring and access demarcation, updated private-cloud software without stopping infrastructure, and performed fault-tolerance tests with the client.

These case studies are not independent audits. They are provider-authored customer stories. Their value is that they reveal the customer problem: the buyer is not asking for an abstract cloud server; it is asking for continuity when a primary data center, storage system, update or human process fails. The server account therefore has to be evaluated by evidence of backup architecture and testing, not by the provider's general brand.

The missing evidence is obvious. Public pages do not disclose median restore time, restore-test frequency, failed-backup rates, ransomware recovery outcomes, number of DRaaS customers, average RPO achieved, or how often customers pay for a recovery configuration but fail to test it. They do not disclose whether a customer buying the lowest-cost OpenStack account gets enough backup discipline to make the "recovery promise" real. A serious buyer would ask for a restore-test report, sample runbook, support escalation path, cross-site replication design, contract liability, and customer references for the same workload class.

Procurement demand shows trust, but not a blank cheque

Public procurement is one of the stronger demand signals for GigaCloud. Opendatabot shows 766 tenders associated with the company and aggregate tender sales by year: 4.916 million UAH in 2021, 92.758 million UAH in 2022, 149.170 million UAH in 2023, 246.352 million UAH in 2024, 483.414 million UAH in 2025 and 232.643 million UAH in 2026 as of the page capture. It lists top public customers by aggregate amount, including DP Diia at 593.830 million UAH, the National Health Service of Ukraine at 166.470 million UAH, the Ukrainian State Center of Radio Frequencies at 76.916 million UAH, Lvivavtodor at 69.880 million UAH, the Ministry of Social Policy at 60.406 million UAH, Infotech at 45.888 million UAH, DP Prozorro at 25.878 million UAH and other public bodies.

Those numbers do not prove service quality, but they do prove something commercially important: public buyers have repeatedly contracted with GigaCloud at meaningful scale. In a market where procurement eligibility, security documentation, local contract form and Ukrainian-language support matter, that is a signal of institutional fit. A foreign hyperscaler can be technically excellent and still be harder for a Ukrainian public user to procure, document and manage under local rules.

The procurement record also creates concentration questions. If DP Diia and other public customers are large contributors to tender sales, the provider may benefit from high-trust public-sector demand but also depend on public budgets, procurement cycles and regulatory eligibility. A change in procurement rules, a failed compliance renewal, a political decision to move workloads elsewhere, a major incident affecting a public service, or a shift toward foreign cloud credits could affect revenue. The public record shows demand; it does not show renewal rates, contract margins, or customer concentration inside private-sector revenue.

The case of Diia is especially important because it connects cloud spending with national digital-service continuity. Opendatabot's aggregate tender sales place DP Diia as GigaCloud's largest named public buyer. GigaCloud's public-sector and NIST pages name eHealth, Prozorro platforms and MilTech-type projects as examples of high-security needs, while the Diia.City United profile says GIGAGROUP serves large state enterprises including Diia and Prozorro. The picture is not of a small web host selling generic VMs. It is of a Ukrainian infrastructure supplier that has positioned itself near public digital systems.

That position raises the standard of evidence. If GigaCloud sells to public and regulated users on recoverability and sovereignty, the most useful public metrics would be incident transparency, service-credit history, security authorization status, public contract performance and backup-test outcomes. Those are not all visible. The public record suggests meaningful buyer trust, but it cannot substitute for operational evidence at the level of the customer's own workload.

Supplier dependence cuts both ways

GigaCloud's product architecture appears to rely on a mix of its own operational work and major technology vendors. Its pages name VMware, OpenStack, Veeam, Lenovo, NVIDIA, Cisco, Microsoft, Fortinet and other partners or technologies. The VMware public cloud page says the virtualization layer is based on VMware technologies and that GigaCloud has Premier Tier VMware Cloud Service Provider status. The NIST page says solutions are based on VMware, uses Lenovo equipment and can be complemented with NVIDIA GPUs. The CSA STAR registry page says GigaCloud cooperates with verified providers such as VMware, Red Hat, Cisco, Lenovo and Microsoft and offers public clouds based on OpenStack or VMware, private clouds, DRaaS, BaaS, VDI and GPU cloud.

This supplier stack gives customers comfort because it uses familiar enterprise platforms. It also creates dependency. VMware's ownership and licensing changes after Broadcom's acquisition have affected many cloud providers globally. Microsoft licensing and security tools are not controlled by GigaCloud. Veeam licensing and product capabilities determine part of the backup economics. Lenovo, NVIDIA and storage-hardware supply chains shape capital expenditure. GigaCloud can package, localize and support the stack, but it does not own all the strategic inputs.

The tariff schedule makes that visible. Windows memory costs more than Linux memory in the OpenStack group. VMware Windows memory costs more than VMware Linux memory. Veeam software lines are priced per virtual machine. Microsoft SQL Server licensing appears as a separate software-access tariff. These are not arbitrary local-provider markups; they reflect the reality that a recoverable server account may include foreign software and hardware economics even when the customer pays a Ukrainian company in UAH.

That matters for the local substitution thesis. Domestic cloud does not mean domestic technology independence. A customer moving from AWS to GigaCloud may reduce dependence on foreign account support, foreign billing and foreign jurisdiction, but it may still depend on VMware, Microsoft, Veeam, Cisco, Lenovo or NVIDIA. The public evidence supports a more modest interpretation: GigaCloud can localize the contract, support, data-center geography and some operating control, but it cannot make global infrastructure inputs disappear.

This is not a fatal weakness. Many enterprises prefer a local provider precisely because someone else manages the vendor stack and absorbs some complexity. But it changes the recovery economics. The buyer should ask whether GigaCloud's supplier relationships, license terms, hardware capacity and spare-parts access are strong enough to support the promised account under stress. The account is worth more if the provider can keep VMware, storage, backup and networking capacity reliable despite wartime logistics and vendor changes.

Hryvnia billing is not a footnote

Currency mismatch is one of the assignment's controlled topics, and it is central to the economics. GigaCloud's English homepage says it is a Ukrainian cloud provider and can accept payments in UAH to a Ukrainian legal entity. The tariff schedule is denominated in hryvnia. The hyperscaler migration page argues that foreign providers' preferential terms for Ukrainian companies could be cancelled and that businesses would then face higher infrastructure costs or need alternatives. It also presents indicative comparisons where foreign cloud options are more expensive than GigaCloud for certain configurations, with the caveat that pricing depends on exchange rates.

The strategic value here is predictability. A Ukrainian company that earns UAH and contracts locally may prefer a domestic provider because the monthly invoice is easier to budget, approve, and reconcile. A public body may also need local procurement documents and VAT treatment. Foreign cloud invoices can still be perfectly manageable for sophisticated teams, but they introduce exchange-rate exposure, foreign tax/accounting routines, card or reseller dependence, and support-tier choices that may be awkward for smaller or public-sector buyers.

The tariff schedule also suggests that GigaCloud is not hiding the fact that service cost is granular. Buyers see VAT, admin hours, storage classes and traffic components. That transparency may be valuable in budgeting even if the total bill is not always lower. A cheaper foreign VM can become expensive if the customer misunderstands egress, support, snapshots, inter-region traffic, storage operations or marketplace licensing. A domestic tariff can also become expensive if the customer adds support hours, remote backup and high-performance storage. The advantage lies in knowing what is being bought and why.

Currency mismatch should not be overstated. GigaCloud itself depends on foreign hardware and software vendors whose costs may be dollar- or euro-linked. The Equinix WA3 option is available only when paid in euros, according to GigaCloud's VMware FAQ. That means GigaCloud cannot fully insulate customers from foreign currency. What it can do, based on the public evidence, is provide a UAH-priced domestic account for many Ukrainian and Ukraine-plus-Warsaw configurations, while reserving euro billing for specific EU-site options.

The business conclusion is therefore conditional. Hryvnia billing adds value when the customer's budget, procurement and revenue are hryvnia-based and when foreign exchange volatility would otherwise create planning risk. It adds less value for exporters earning in euros or dollars, global software companies already operating on hyperscalers, or workloads where foreign region depth and service breadth are worth the currency exposure.

Competitors are not only other Ukrainian clouds

GigaCloud competes with several kinds of substitutes. The obvious foreign substitutes are AWS, Microsoft Azure, Google Cloud and Oracle Cloud. They offer global scale, broad managed services, deep automation, multi-region patterns, mature observability and large partner ecosystems. For many software companies, they remain the default because engineers know them and because platform services reduce application-development burden.

Ukrainian and regional infrastructure competitors are also relevant. De Novo, Parkovy, Tet, colocated private hardware, telecom-linked data-center services, managed hosting firms and internal IT departments can all compete for the same server account. A customer may decide that a colocation rack plus backup to Warsaw is more controllable than a cloud account. Another may use a local cloud for public-sector systems and a foreign hyperscaler for product analytics or customer-facing international services. The market is not binary.

GigaCloud's differentiation is strongest where the account needs a combination of domestic contract, public-sector documentation, local support, VMware familiarity, Ukraine/EU data-location options and backup/DR. It is weaker where the workload needs advanced cloud-native services, global edge, proprietary managed databases, large AI platform services or engineering teams already optimized around foreign hyperscalers. GigaCloud offers GPU cloud and AI-related products, but the public evidence in this assignment is stronger for enterprise IaaS and continuity than for a full hyperscaler-equivalent platform.

The customer dependence problem also differs by buyer type. A public agency choosing GigaCloud may gain procurement and compliance fit but may lock into provider-specific support and recovery patterns. A private company leaving AWS may reduce foreign billing exposure but may lose access to certain managed services or global automation. A company leaving self-hosted servers may reduce capital expenditure and internal labor but may become dependent on a provider's incident handling and contract renewal terms.

The best reading of the public evidence is that GigaCloud is not trying to beat hyperscalers on every dimension. It is selling an account for workloads where recoverability, local support, local documentation and geography matter enough to overcome the benefits of global commodity cloud. That is a defensible position, but only for workloads whose risk model matches it.

Technical public records show a real surface, not internal resilience

Public technical records add context but should not carry the main conclusion. RIPE's member list places Gigacloud LLC in Ukraine with a Kyiv address. BGP.tools lists AS49720 for Gigacloud LLC, with originated IPv4 and IPv6 prefixes and rankings in Ukraine. DNS records for gigacloud.ua show GigaCloud-controlled nameservers, Microsoft 365 mail protection, Microsoft verification, Google site verification, Duo SSO verification, Atlassian domain verification, DKIM and SPF including Microsoft, SendPulse, Selzy and several IPv4 addresses.

Those records show that GigaCloud has a public internet and SaaS operating surface consistent with an active cloud provider. The Microsoft mail-protection and verification records suggest Microsoft 365 usage; Duo and Atlassian verification suggest identity and collaboration-service integrations; SPF entries show named email and marketing providers plus local IPs. None of this proves customer data residency, internal security posture, backup design, data-center uptime or incident maturity. It is supporting evidence, not decisive evidence.

The public technical footprint also illustrates a broader point: even a domestic cloud provider uses global SaaS and software vendors for its own corporate operations. Domestic cloud substitution is not isolation. It is selective localization of customer infrastructure, support and contract control within a global technology stack.

For a buyer, the technical due diligence should go deeper than public DNS. It should include architecture diagrams, network path diversity, peering and transit arrangements, DDoS protection, RPKI and routing practices, backup isolation, identity and key management, administrative access controls, audit logs, support escalation, maintenance windows, power-resilience documentation and proof of restore testing. The public records justify asking those questions; they do not answer them.

Power, network and war risk are the price of the Ukrainian promise

The recovery promise cannot ignore Ukraine's operating environment. A domestic cloud account may reduce some risks and increase others. It can reduce jurisdictional uncertainty, procurement friction, support friction and latency to Ukrainian users. It may increase exposure to physical attacks, power disruption, mobilization pressures, local network incidents, wartime logistics and sudden public-sector demand spikes.

GigaCloud's public materials acknowledge the need for multiple sites and business continuity. The Lviv expansion story is framed around preparing for force majeure and possible attack on Kyiv. The backup and replication page emphasizes RTO and RPO. The official product pages mention five technical sites in Ukraine and the EU. This architecture is the right answer in principle: do not depend on one city, one data center, one storage system or one restore path.

The public record does not show enough to quantify power and network resilience. It does not disclose generator runtime, fuel contracts, battery capacity, power-feed diversity, actual wartime outage minutes, network transit concentration, customer-impacting incidents or restoration times by site. It also does not show how many customers buy cross-site redundancy rather than a low-cost single-site server.

This is where the customer's purchasing discipline matters. A GigaCloud account bought only as cheap local compute is exposed to the same old failure pattern: a server exists, a backup may exist, but no one knows whether the service can be restored under stress. A GigaCloud account bought as a resilience contract should specify site placement, backup frequency, restore point, restore time, support escalation, testing cadence and contract liability. The provider can offer the components, but the buyer has to pay for the recovery design.

Public evidence worth reading

The public record is strongest when sources are read together rather than in isolation.

GigaCloud's VMware public cloud page, https://gigacloud.ua/en/products/public-cloud-vmware/, supports the claim that the company sells VMware-based IaaS with monthly billing, 99.95 percent SLA for that product, Kyiv/Lviv/Warsaw data-center options, certifications, backup and replication add-ons, and EU hosting options including Atman and Equinix WA3.

GigaCloud's backup and replication page, https://gigacloud.ua/en/solutions/rezervuvannya-danih-u-hmari/, supports the claim that the company sells BaaS and DRaaS based on Veeam and VMware, frames the product around RTO and RPO, and builds continuity services across Ukraine and the EU.

GigaCloud's tariff PDF, https://gigacloud.ua/wp-content/uploads/2024/12/taryfy-na-poslugy-vid-1-zhovtnya-2025-1.pdf, supports the account-level economics: resource, storage, connectivity, IP, backup, Veeam, Microsoft and admin-hour tariffs in hryvnia effective October 1, 2025.

Opendatabot's GigaCloud company page, https://opendatabot.ua/c/39792589, supports the legal identity, Kyiv address, company code, registration date, ownership entries, financial indicators, employee count and public-procurement sales signals.

RIPE NCC's member page, https://www.ripe.net/membership/member-support/list-of-members/ua/gigacloud/, supports the fact that Gigacloud LLC appears as a RIPE member in Ukraine with a Kyiv address. It should be used only for public-infrastructure context.

BGP.tools, https://bgp.tools/as/49720, supports the public routing surface associated with AS49720 for Gigacloud LLC. It cannot prove internal architecture or service quality.

GigaCloud's NIST-authorized cloud page, https://gigacloud.ua/en/products/cloud-with-nist/, supports the public-sector compliance claim that its cloud infrastructure is authorized under Ukraine's State Service of Special Communications and Information Protection requirements and listed in the relevant official registers.

GigaCloud's sovereign cloud page, https://gigacloud.ua/en/solutions/sovereign-cloud/, supports the company's data-sovereignty positioning, claims about data storage and processing within national borders, and inclusion in the SSSCIP list.

Asters' note on new Ukrainian cloud and data-center regulation, https://www.asterslaw.com/press_center/publications/new_regulation_of_cloud_and_data_center_services_in_ukraine/, supports the regulatory context: service-provider requirements, official lists, electronic catalogues, model contracts, incident notification and public-user eligibility after the transition period.

Lviv IT Cluster's account of GigaCloud's Lviv expansion, https://itcluster.lviv.ua/en/gigacloud-cloud-operator-launches-an-updated-data-centre-in-lviv-why-it-is-important-for-business-in-the-context-of-war/, supports the war-time resilience context, including Lviv technical capacity, backup storage and network independence claims.

GigaCloud's Prozorro case, https://gigacloud.eu/cases/private-cloud-for-reliable-operation-and-fault-tolerant-system-prozorro-case/, supports the market signal that GigaCloud has sold private-cloud and fault-tolerance work for Ukraine's public procurement platform, while remaining a provider-authored case study.

CSA STAR's GigaCloud registry page, https://cloudsecurityalliance.org/star/registry/gigacloud, supports the third-party registry context for security self-assessment and the provider's stated offerings, certifications and 99.95 percent SLA claim. It is not a performance audit.

Diia.City United's profile of Nazariy Kurochko, https://diiacityunited.org/en/diia-city-united-welcomes-two-tech-visioners-to-its-strategic-board/, supports group context for GIGAGROUP and its businesses, including GigaTrans, GigaCenter, GigaCloud and GigaSafe.

GigaCloud's hyperscaler migration page, https://gigacloud.ua/services/hyperscalers-migration/, supports the currency and substitution argument. Because it is a sales page, its relative-cost comparisons should be treated as a provider claim rather than independent market pricing.

What would change the judgment

Several facts would materially change the conclusion. A public incident history showing repeated customer-impacting failures without transparent remediation would weaken the recovery promise. Published restore-test metrics, customer-level DRaaS outcomes, actual SLA credit history or third-party audits of backup controls would strengthen it. Clear disclosure of customer concentration, public-sector contract margins and renewal rates would clarify demand quality. Data-center power-resilience details, network transit diversity and cross-site replication performance would show whether the geography is operationally real. A detailed comparison of GigaCloud bills against AWS, Azure, Google Cloud, Oracle Cloud and Ukrainian competitors for the same workload would test the price argument properly.

The buyer should also watch vendor dependence. VMware licensing changes, Veeam pricing, Microsoft licensing, NVIDIA GPU availability, Lenovo hardware supply and Equinix/Atman site costs could all change the account economics. The fact that GigaCloud invoices many services in UAH does not mean its underlying cost base is immune to global vendor prices. If foreign software or hardware costs rise sharply, local tariffs may follow.

Finally, the buyer should distinguish between provider capability and purchased configuration. GigaCloud may have five technical sites, but a customer who buys only one-site compute does not have five-site resilience. GigaCloud may offer DRaaS, but a customer that never performs a restore test has bought hope, not recovery. GigaCloud may be listed for public procurement and security authorization, but a private workload still needs its own access controls, backup policy and incident process.

The evidence supports a conditional recovery premium

The evidence supports the core thesis with limits. GigaCloud's cloud server account is valuable when the customer treats it as a recovery contract: a priced bundle of compute, storage, connectivity, backup, replication, support, data-location choice and compliance documentation. The public record suggests that GigaCloud is a credible Ukrainian provider for that use case. It has legal identity, revenue scale, public-procurement demand, published tariffs, Ukraine/EU site claims, backup and DR products, RIPE membership, public routing evidence, security registry claims and customer cases tied to continuity.

The same evidence does not support a blanket claim that domestic hosting is always safer or cheaper than foreign capacity. A foreign hyperscaler may be better for global reach, managed platform breadth, advanced automation, cross-region scale, or teams already built around those tools. Another Ukrainian data-center or cloud provider may be better for a specific workload, location or contract. An in-house server may be rational where the workload is stable, low-risk and already well managed.

The public record suggests that GigaCloud is worth paying for when the avoided failure cost is high and when local factors materially reduce that cost: Ukrainian support, UAH invoicing, public-sector procurement eligibility, data-location control, Kyiv/Lviv/Warsaw placement, Veeam-based backup, DRaaS and documented compliance. The thesis remains unproven without customer-level metrics for actual uptime, restore tests, incident response, service credits, churn, workload pricing and cross-site recovery performance. Until those metrics are public, the right conclusion is not that GigaCloud's server bill is cheap. It is that, for the right Ukrainian buyer, the bill can be a rational recovery premium rather than a patriotic surcharge.