Automating Debt Recovery Before the Law Required It: A Conversation with Botagoz Karimova

Named Employee of the Year at a major Kazakhstan bank, she built collection automation systems years before the country’s 2025 banking reform made them mandatory. 

Recent regulatory changes in Kazakhstan’s banking sector have introduced new requirements for data integration and monitoring processes. Among the most ambitious new requirements, according to Baker McKenzie’s legal analysis, all banks must integrate with a unified “Data Showcase” information system by January 1, 2027, and introduce biometric identification for online lending. For an industry where many collection and monitoring processes still rely on manual workflows, the deadline is tight.

Botagoz Karimova spent years solving exactly this kind of problem, effectively addressing challenges that later became part of broader regulatory changes. As a debt recovery specialist at one of Kazakhstan’s largest retail commercial banks, serving millions of clients nationwide, she designed and launched an automated pre-judicial collection system that connected internal workflows to government databases, notary platforms, and external enforcement partners. Her work earned her the Employee of the Year award in 2023 and a recommendation letter from the bank’s vice president. Now based in Portland, Oregon, where she works as a Loan Support Specialist at a credit union, Karimova sat down to explain how she built the automation infrastructure from scratch, what her mediation pilot achieved for hundreds of borrowers, and what the transition from Kazakh banking to the American credit union model taught her about the limits and possibilities of digital finance.

Botagoz, the 2025 amendments to Kazakhstan’s banking law require all lenders to integrate with a centralized digital monitoring system by 2027. At your bank, you automated an eight-stage collateral realization workflow, connected it to the national notary platform and a government enforcement database, and set up daily reconciliation across all of them, effectively anticipating these requirements years in advance. Where did that effort begin?  

When I joined the department handling problem loans, nearly every step of the pre-judicial collection process was done manually. Staff tracked notary enforcement orders in spreadsheets, communicated with bailiffs by phone, and compiled reports by hand. Nobody had a clear picture of how many cases were active, at what stage, or whether deadlines were being met. I started by mapping out each workflow, identifying where the biggest time losses occurred, and then worked with the IT team to automate status tracking and deadline alerts. Connecting to the government’s enforcement proceedings database was the first major milestone, because it gave us a real-time view of case statuses instead of waiting for paper updates. After that, the integration with the national notary platform let us generate enforcement documents digitally, which cut processing time significantly. None of it happened as one big rollout; we solved one bottleneck, proved it worked, and moved to the next.

Integration with government platforms sounds straightforward on paper, but government systems can be complex to integrate due to differences in processes and data formats. What made this particular integration so challenging?  

Nothing about government systems is straightforward. Approval cycles are long, data formats don’t always match what banks need, and any change on the government side can break an integration overnight. We had to design our monitoring dashboards in Power BI to accommodate inconsistencies in the data we received, and there were no ready-made solutions on the market, so many processes had to be built from scratch. On top of that, regulatory requirements continue to evolve, so systems need to remain flexible and adaptable. What helped most was having direct relationships with external partners such as notaries, private bailiffs, and collection agencies. I could understand their constraints and design the system around real behavior, not theoretical processes.

One of the more unusual projects in your track record is the mediation pilot that led to writing off 450 problem loans worth over 310 million tenge – roughly $626,000. How did that initiative come together?  

Mediation was not a common tool in Kazakhstan’s banking practice at that time, especially for retail problem debt. I proposed the pilot because we had a segment of loans where traditional collection methods like enforcement and collateral seizure were either too expensive or unlikely to recover meaningful amounts. For the borrowers, these debts had become a permanent burden with no realistic path to repayment. By structuring mediation agreements that satisfied both the bank’s write-off criteria and the borrowers’ ability to settle, we managed to clear 450 cases from the portfolio. It reduced the non-performing loan volume, freed up staff resources, and gave borrowers a way out. Not every bank was willing to try this approach, but the numbers made the case for us.

Beyond your bank, you’ve also been involved in shaping how future finance professionals are trained. In November 2025, Almaty Technological University invited you as an external expert to review their bachelor’s curriculum in accounting and finance. What did you find needed changing?  

University programs in Kazakhstan, and honestly in many countries, tend to lag what the industry actually needs. Students graduate knowing accounting theory and regulatory frameworks on paper, but they’ve never built a dashboard in Power BI, never calculated a borrower’s debt-to-income ratio using real data, or analyzed a profit-and-loss statement from a functioning business. My recommendations focused on practical skills: expanding Excel training beyond basic formulas to include pivot tables and financial reporting templates, introducing Power BI for data visualization, and adding credit analysis exercises that mirror real-world lending decisions. I also pushed for a module on financial data protection, because anyone handling client information in a bank needs to understand the sensitivity of what they’re working with. Almaty Technological University accepted these recommendations and included them in the official protocol for the 2026 academic year.

You’ve recently transitioned to the US financial sector, working at a credit union in Portland, Oregon. How does American lending compare to what you experienced in Kazakhstan?  

Every financial system has its own logic. In Kazakhstan, I worked within a centralized banking model where one institution handles everything from origination to collection to write-off. Credit unions in the US operate on a completely different philosophy – they’re member-owned, community-oriented, and the regulatory environment is built around consumer protection in ways that Kazakhstan is only beginning to adopt with its new law. What stays constant, though, is the underlying analytical work: assessing creditworthiness, managing risk, ensuring compliance. I speak Russian, Kazakh, English, and Chinese, which I studied in Xi’an, Shaanxi province. That range helps me communicate across cultural contexts, and it matters more than people might expect in a country as diverse as the United States.

After spending years building systems from scratch in an environment with few ready-made solutions, what advice would you offer a young specialist entering banking technology today?  

Learn the business before you learn the tools. Automation fails when it’s built by people who understand software but not the process it’s supposed to improve. I succeeded not because I had the best technology, but because I had spent years doing the manual work myself – I knew which steps were bottlenecks, which reports nobody read, which handoffs between departments created delays. If you understand the pain, you can design the solution. Otherwise, you’ll build something elegant that nobody uses.

Have you read?
Europe’s Golden Visa Split: Spain Tightens, Hungary Opens.
Italy’s Golden Visa Surge: A New Era of Lifestyle Investing.
Rustam Gilfanov: Bridging Research, Manufacturing, and AI.
What Swiss Industrial Families Know About Capitalism.
From Audit to Execution: Bakhо Muratov’s Strategy.