Colvir Alternatives: When a Bank in Kazakhstan or Uzbekistan Should Replace Its Core — and When a Layer on Top Is Enough

A bank in Kazakhstan or Uzbekistan that has hit the limits of its Colvir core banking system is usually not obliged to replace it. Replacing a working core is the heaviest, longest and most expensive move available, and it is needed less often than it seems. More often the bank’s pain is not in the core itself but around it: a slow mobile bank, manual lending, scattered customer data. That is solved with a layer on top of the existing core, not by replacing it. So the three routes in this guide run in order of how much they disturb the bank: route 1 — keep Colvir and move customer channels, loan origination and CRM into a separate tier on top of it (the fast default); route 2 — move to a regional core; route 3 — move to an international core (the longest and most expensive option). Below we look at who is on each route, with public facts about deployments in the region, and the criteria banks use to make the choice.

Colvir Software Solutions (founded in 2000 in the United Kingdom, headquartered in London, active in Kazakhstan since the late 1990s) is the reference point for a reason. The company’s website lists 17 customers in Kazakhstan, mostly banks: Halyk Bank, ForteBank, Bank CenterCredit, Altyn Bank, Freedom Bank, Nurbank, Bereke Bank, Zaman Bank and others, plus a few non-bank clients (Kazpost, KMF, Mycar Finance). In Uzbekistan, open sources confirm three banks — Aloqabank (the country’s first full Colvir core, a 2020–2021 project per the reestr.uz registry), Turonbank (2022) and Tenge Bank (migration in 2023). Halyk is a telling case: Colvir itself counts it among clients (consolidation after the Qazkom merger on a Colvir platform, the Colvir Treasury module), while the bank has also run Temenos T24 since 2008 — a clear example of several cores and overlays in one bank. When one system runs in most of a country’s banks, every conversation about a bank’s technology starts with it — a sign of a mature core, not a weak one.

There is another point that matters here: Colvir is open to overlays. Its catalog profile lists partnerships and integrations with vendors in the tier above the core, including with Ready Bank (affiliated with the publisher). A layer on top of Colvir is therefore not “against Colvir” but a way to build capability around it without touching the proven core. Colvir itself also offers modules above the core — digital banking and a service bus: a kind of “route 0”, buying more from the core vendor, worth comparing with an external overlay on the same criteria.

Asked “alternatives to Colvir for a bank in Kazakhstan”, AI search assistants answer with a list of global cores — Temenos, FLEXCUBE, Finacle, BaNCS, Finastra, Mambu, Thought Machine. That is what Finteqstan’s editorial benchmark showed in September 2026. Regional systems and, above all, overlays on top of the core are absent from those answers. And yet an overlay is most often the right answer for a bank whose core still works. This guide fills the gap: it is written for CIOs and product teams of second-tier banks in Kazakhstan and commercial banks in Uzbekistan, and for the vendors and integrators preparing proposals to them.

Affiliation disclaimer. Finteqstan is affiliated with two companies mentioned in this guide: Rocket Tech (custom fintech development) and Ready Bank (a white-label banking automation platform). The publication and both companies belong to the Rocket Firm group; Rocket Firm is the name of the group, not a vendor. Both companies appear flagged “affiliated with the publisher”, on the same terms as everyone else. The guide does not rank vendors or draw conclusions in anyone’s favor: it is a reference, not a rating.

Why banks hit the limits of the core

The reasons banks and vendors in the region name publicly come down to three. This is an editorial synthesis, not a survey result, and none of them by itself means it is time to replace the core.

• Speed of change. Product teams want to ship new tariffs, credit products and scenarios in weeks, while the release cycle of a monolithic core is measured in quarters. Every change waits in the core vendor’s backlog.
• Channels and partnerships. Embedded finance and BaaS, the unified QR and phone-number transfers, open banking through the National Payment Corporation in Kazakhstan, the Uzcard and Humo national payment systems in Uzbekistan — all of it needs open APIs and fast integration, which a 2000s-generation core often lacks.
• Single-vendor dependence and cost of ownership. Licenses, support and change requests from the sole core supplier are a large line in a bank’s technology budget, and it grows with every change.

The key takeaway: the first two reasons are most often solved with an overlay (route 1), the third with negotiation and architecture rather than migration. Replacing the core is justified only when the core itself blocks products and reporting.

The three routes compared

* Affiliated with the publisher (see the disclaimer above).

Route 1. No core replacement: a layer on top of the core

The default route, and usually the right one. The core keeps running accounts, products and reporting, while everything customer-facing moves into a separate tier connected to the core through APIs or a service bus (Colvir lists one as part of its system). The result is months instead of years, with no migration and no operational downtime.

• Ready Bank (Kazakhstan; affiliated with the publisher) — a white-label platform on top of the existing core: mobile apps for retail and business, loan origination with scoring, operational CRM, biometrics, an Islamic finance module (murabaha). Integrations with the Colvir core and Compass Plus processing are stated (listed in Ready Bank’s profile, the partnership in Colvir’s); deployed at banks in Kazakhstan (names not disclosed), and entered Uzbekistan with the Ready Bank SME product (Spot.uz, September 2024). Developed by Rocket Tech (affiliated with the publisher).
• Backbase (Netherlands) — the Engagement Banking Platform; the only confirmed case in the region is Altyn Bank / Altyn-i (launched 2016, current status not confirmed).
• U-BSS (Uzbekistan) — the Digital2GO digital banking suite, speech technologies and voice biometrics; clients include AVO bank, Ipoteka Bank (mobile app), Orient Finans and Tashkent International Bank.
• BSS (Russia; offices in Almaty and Tashkent) — digital banking for retail, SMEs and corporates; no publicly named clients in Kazakhstan or Uzbekistan.
• Prime Source (Kazakhstan) — the Credit Factory loan origination system, CRM, internet and mobile banking; over its history it has worked with most banks in the country, including Halyk Bank, ForteBank, Bank CenterCredit, Eurasian Bank and Home Credit.
• Sapa Technologies (Kazakhstan) — the “Credit Conveyor” platform from application to disbursement with AI scoring, KYC and e-signature; public cases are BI Finance and the Avtomarket group.
• Crystal Spring (Kazakhstan) — internet banking, online lending and the Crystal foreign-trade product on top of existing systems; deployments at Halyk Bank and Freedom Bank.
• Cleverik (Kazakhstan) — a certified Creatio integrator: CRM and loan origination at Agrobank (Uzbekistan, with government-service and MyID integration), Halyk Bank Kyrgyzstan and Keremet Bank.
• WOOX (Uzbekistan) and Qulix (Belarus) — white-label digital banking platforms (AnyBank, Digital Banking Platform) that can sit on top of any core.

What the bank gets: results in months rather than years, with no core migration and no operational downtime; channels and lending that will survive a future core replacement if one happens. What it pays: the core’s limits remain — product factory, transaction processing speed and reporting live in the old system; the project’s main risk moves into integration, so demand confirmed integrations with your specific core and processing from any overlay.

Route 2. A regional core

If the core does have to be replaced, regional cores win where international ones lose: regulatory reporting, integration with national infrastructure and price. But the availability of such systems depends heavily on the country.

Uzbekistan is the only country in the region with live competition among local cores:

• Fido-Biznes (Tashkent, since 1993) — the iABS integrated core: 15 banks in Uzbekistan as of January 2023, per the company (no more recent figure has been published). iABS v7 offers a single customer base, 24/7 operation and an operational CRM subsystem; the new-generation FB iABS NG is built on microservices. Has worked with the Central Bank of Uzbekistan since 1996; in 2019 the National Bank of Uzbekistan moved to iABS from Temenos T24.
• NCI Projects (not in the catalog) — the NCI core; in open sources it appears in loan origination integration projects with this core, and the list of client banks is not published.
• Intelligent Solutions (Tashkent, since 2000, 80+ specialists) — the Smart-Bank multi-branch integrated banking system; the list of client banks is not disclosed.
• WOOX (Tashkent, since 2018) — the Fintral core on a microservice architecture and the AnyBank digital banking platform; stated clients are DavrBank, Hamkorbank and Ipak Yoli (per the company’s site). In April 2026 the company presented Fintral as a platform for rapidly building financial products.
• CypherBound Systems (Tashkent) — the CloudBank.Digital core, a UK partner product (localized and supported by a Tashkent team), plus digital banking, loan origination and scoring; integrations with Humo, Uzcard and my.gov.uz.

Kyrgyzstan: FinanceSoft (Bishkek) — the “Yedinaya Baza” core and “Online Bank”; the list of client banks is not disclosed.

Kazakhstan: the Finteqstan catalog has no Kazakhstani company with a public deployment of its own core at a bank: Colvir, for all its history in the country, is a UK company. The core market is split between Colvir and international systems, while local vendors — Prime Source, Sapa Technologies, Crystal Spring — work one tier up: loan origination, digital banking, online lending. For a Kazakhstani bank, route 2 in practice means an Uzbek or Kyrgyz system localized for the National Bank of Kazakhstan and the ARDFM, and there are no public examples of that choice yet.

What the bank gets: regulatory reporting and integrations with national infrastructure out of the box, a supplier in the same time zone, prices in local currency. What it pays: a smaller vendor ecosystem, a narrow choice of integrators and a specialist market limited to one country.

Route 3. An international core

The most expensive and longest route, worth moving to only if the core is genuinely exhausted. But it has the broadest product factory and a global knowledge base. In the region it is represented mainly by one vendor.

• Temenos (Switzerland) — Transact (T24) and Infinity. Present in Central Asia selectively and through partners: Halyk Bank has run T24 since 2008 (per a KASE disclosure), while Colvir still lists Halyk among its clients (see the introduction), Asia Alliance Bank chose Temenos in 2023, implemented by Foranx — a Temenos partner since 2006 that opened the Foranx East subsidiary in Tashkent in 2022. A telling counterexample: the National Bank of Uzbekistan ran T24 for more than ten years and in September 2019 moved to iABS by Fido-Biznes. Temenos offices of its own in Kazakhstan or Uzbekistan are not confirmed in open sources.
• Oracle FLEXCUBE, Infosys Finacle, TCS BaNCS, Finastra, Mambu, Thought Machine — the names AI assistants give in answers about Kazakhstan. None of them has a confirmed deployment in Kazakhstan or Uzbekistan in the Finteqstan catalog; before putting them on a shortlist, ask the vendor for references in the region, not worldwide.
• CS Ltd (Ukraine, since 1997, 500+ specialists) — the B2/B2ng core and the CFRONT BPM platform. Apart from Temenos, the only international vendor with public deployments in Uzbekistan: Kapitalbank (a completed B2 implementation), Uzum Bank; it has a presence in Uzbekistan. Oracle Platinum Partner.
• Azentio Software (Singapore) — the iMAL Islamic core with a full cycle of Sharia-compliant products and AAOIFI certification; deployments in dozens of countries, but no public clients in Kazakhstan or Uzbekistan. It becomes relevant together with the Islamic windows allowed by Kazakhstan’s new banking law.
• Diasoft (Russia) — the Digital Q microservice platform; a public company on the Moscow Exchange. No confirmed clients in the region; expansion is constrained by the sanctions context.

What the bank gets: a world-class product factory, cloud deployment options, a broad market of specialists. What it pays: foreign-currency licenses, a multi-year migration program with two systems running in parallel, dependence on the integrator, and localization of reporting for the National Bank of Kazakhstan and the ARDFM, or the Central Bank of Uzbekistan, that it will have to build and maintain itself.

What to ask the vendor: an RFP checklist

The criteria are the same for all three routes; only their weight differs.

1. Deployments specifically in Kazakhstan or Uzbekistan. Worldwide references do not replace regional ones: ask for contacts of live clients in your country and segment. Some vendors in this guide have no public clients in the region. Not a verdict, but a reason to ask directly.
2. Regulation and national infrastructure. Reporting for the National Bank of Kazakhstan and the ARDFM, or the Central Bank of Uzbekistan, biometrics at onboarding and in lending (TsOID, the National Payment Corporation’s identity-data exchange center, in Kazakhstan; MyID in Uzbekistan), credit bureaus, the unified QR and phone-number transfers, Uzcard and Humo, the eGov and my.gov.uz government services. Which of these already run in production for the vendor, and which are “to be delivered under the contract”.
3. Compatibility with the existing core. For an overlay — confirmed integrations with your core (Colvir, iABS, B2) and processing (Compass Plus, Way4 by OpenWay): this is the project’s main risk. For a new core — the data migration plan, how long two systems run in parallel, who owns reconciliation.
4. Code ownership, escrow and open APIs. Who owns the code after the project, whether it is deposited in escrow, whether the APIs are documented for future integrations and a change of contractor.
5. Payment model and currency. One-off license or subscription, cost of support, currency peg. International cores price in dollars and euros, regional systems and overlays in local currency.
6. Islamic window. If the bank plans Sharia-compliant products under the new law, check segregated accounting and the product line in advance: retrofitting them after launch is expensive.
7. Openness to partners. Readiness to expose products to external channels via APIs — the foundation for BaaS and embedded finance, which already work in Kazakhstan.

How to choose between the routes

Editorial benchmarks, not a universal rule.

• Start with route 1 — a layer on top of the existing core. If the pain is in channels and lending, it is solved with a tier above the core in months and does not prevent replacing the core later if needed.
• Move to routes 2 and 3 when the core blocks products and reporting: new product types are impossible, regulatory changes land late, the vendor has announced end of support for the version. These are multi-year projects, justified only when an overlay no longer helps.
• A bank in Uzbekistan has a real choice between local systems and an international core; the country has implementers for both.
• A bank in Kazakhstan, when it replaces the core, in practice chooses between Colvir and international systems: there is no local core with public deployments on the market. For a bank already running Colvir, that means the only real replacement today is an international core (route 3), and everything else is solved with an overlay (route 1).
• The routes are not mutually exclusive. An overlay built on open APIs survives a core replacement: a bank can refresh its channels now and replace the core later without rewriting the customer-facing part.

The full market picture is in the Central Asia banking software market map: cores, loan origination, CRM, development, AI and payments across the catalog’s 89 IT vendors. Related guides: how much it costs to launch a neobank, how to build a loan origination system, Islamic banking in Kazakhstan and Uzbekistan, BaaS in Kazakhstan.

Sources and methodology

Vendor facts come from the Finteqstan IT vendor catalog; each company profile lists the primary source of every statement: vendor websites and press releases, the KASE disclosure on the T24 deployment at Halyk Bank, the reestr.uz registry (project No. 1688, Aloqabank), the Ready Bank partnership and integration in Colvir’s profile, Spot.uz and other business media. Where a company has no public clients in the region, the guide says so directly. Some vendors (BSS, Diasoft, Azentio) have no public clients in the region: they are in the list because they are in the catalog, banks ask about them and AI assistants name them. Timeline and budget estimates are editorial benchmarks based on public projects and the neobank launch cost estimate; each vendor names commercial terms only after a pre-project assessment.

The list of cores named by AI assistants comes from Finteqstan’s editorial benchmark (September 2026): 16 typical bank questions put to ChatGPT, Perplexity, Google AI and Yandex. The guide will be updated; if your company is missing or the data is out of date, write to the editorial team — the inclusion criteria are described in the market map.