Digital Business, together with Astana Hub, is launching the project «100 Startup Stories of Central Asia». The initiative will spotlight promising projects from Kazakhstan, Uzbekistan, Kyrgyzstan, and Tajikistan with the potential to become future unicorns. Featured companies will include established IT firms that have already achieved product–market fit and begun international expansion, as well as young startups with strong growth potential. The project will be published in several languages, including Kazakh, Russian, English, Kyrgyz, Uzbek, and Tajik.
We begin the «100 Startup Stories of Central Asia» series with the story of Altbridge, founded by Kazakhstani entrepreneur Nazym Azimbayev and Andrew Kim. The two met while studying mathematics at Moscow State University, but their paths later diverged. Andrew became a cryptographer and moved to South Korea, while Nazym spent many years working at the National Investment Corporation of the National Bank of Kazakhstan. Last year, the friends reunited to launch their startup. Their team is developing an AI agent that collects and analyzes financial data to help investors more quickly identify promising companies for investment.
In the interview, Nazym and Andrew shared how their passion for mathematics led to the creation of their own project, the original idea behind Altbridge, and the reasons they had to transform the product. They also explained what gives their solution an edge over other AI agents and how a world-renowned investment expert joined their team.
«We held biweekly calls to review the university math curriculum»
– Nazym, Andrew, what were each of you doing before founding Altbridge?
Nazym: – I graduated from the Faculty of Applied Mathematics at Moscow State University. However, from the very beginning, my career was closely tied to investments. Before Altbridge, I worked for many years in the Department of Monetary Operations at the National Bank of Kazakhstan as a Senior Portfolio Manager. I was responsible for managing the global bond investments of the National Fund’s gold and foreign exchange assets, similar to a Global Fixed Income strategy (an investment approach focused on debt instruments worldwide, noted by Digital Business).
Nazym Azimbayev
Later, I briefly joined KASE and then became CIO at the National Investment Corporation of the National Bank. I was responsible for the strategic allocation of assets across various classes, including private equity, real estate, and public markets. A particular focus was on hedge funds (investment vehicles that pool investors’ capital and allocate it with the aim of maximizing returns while minimizing risk, noted by Digital Business).
After that, I became involved in the digital transformation of the National Bank. It became clear that the financial and investment sectors are undergoing rapid change, and technologies, including large language models, can help institutions respond to these shifts more effectively.
Andrew: – I am also a mathematician. In high school, I won a gold medal at the International Mathematical Olympiad (IMO). I then entered the Faculty of Mechanics and Mathematics at Moscow State University and later earned a PhD in Cryptography at Seoul National University.
Andrew Kim
For about 10 years, I worked with homomorphic encryption, a technology that makes it possible to perform various operations on data without decrypting it, such as multiplication, division, and addition. All of my academic research was dedicated to this field. Before co-founding Altbridge, I worked at Samsung Electronics.
– What inspired the idea behind Altbridge?
Nazym: – I’ve always enjoyed creating my own projects. In my fourth and fifth years at university, I launched two startups and participated in accelerators in Russia and Finland.
The idea for Altbridge came about in an unusual way. Andrew and I first connected during our studies at MSU, but after graduation, our paths diverged: I moved to Astana, and Andrew to Seoul. At some point, we decided to refresh our math knowledge. We set up two time slots and held regular calls, where we would take turns teaching each other parts of the curriculum. Whoever missed a session had to pay a $100 penalty. We kept this up for two and a half years. Through these meetings, we realized that mathematics could be applied to investments, and that’s when we began thinking about how to turn this into a real project.
«The AI shows whether a particular investment suits the client»
– What did you start with?
Nazym: – The first prototype was created to meet my own needs. I enjoy listening to investment-related podcasts, but at some point, there were so many that I simply couldn’t keep up. That’s when Andrew and I built an AI agent that could go through the podcasts and extract all the key insights.
We realized the solution worked well and decided to test whether anyone would actually buy it. This was last summer, just as the market was beginning to explore how AI could be applied in investments.
Andrew: – In addition to podcast transcription, we developed a proprietary relevancy model (an algorithm that determines whether an object, document, message, or search result is relevant to a specific request, context, or user intent, note by Digital Business). We began applying this model in our in-depth research projects, based on the investment criteria set by each client.
Within 3–4 hours, our AI agent could analyze more than 150 different sources, ranging from news articles, scientific papers, and journalistic materials to social media posts, and generate a 20–30 page report. This document included all available information from the web, future forecasts, and actionable recommendations. By comparison, a professional analyst would typically spend 80–90 hours producing the same result.
– But the idea eventually changed. What made you decide to pivot the product?
Andrew: – We began attracting our first paying clients, and things were going quite well. But soon we realized that major players such as OpenAI and Google were rapidly catching up with us in terms of technology. Before long, solutions like Deep Research appeared from nearly all of the big labs (OpenAI, Google, Anthropic). Although their products were still weaker than ours and clients continued to subscribe to Altbridge, it became clear that we would not survive the next wave of competition.
That’s when we started thinking about how to focus on more specific needs and remain valuable to our clients.
– So what is your product offering now?
Nazym: – We continue to provide in-depth reports for some of our clients. At the same time, we are developing an AI agent for investment research and gradually onboarding clients to the platform. Its principle is similar to Spotify: while the streaming service recommends tracks based on a user’s listening history, Altbridge analyzes a client’s typical investment activity and suggests the most relevant investment opportunities.
– How does it work?
Nazym: – An investment profile is created for each client. The system analyzes user behavior, and our AI agents fine-tune themselves to it, enabling them to recommend positions that should be added to the client’s dashboard.
Each investment position has specific drivers that influence its price, such as news, company reports, and market trends. Taking these factors into account, along with the client’s individual profile, the system calculates an overall score that shows how suitable a given investment is for that person.
The AI agent draws on a wide range of sources, including scientific articles, analytical reports, and expert posts on social media. In essence, it identifies information that could potentially affect the investment accuracy of each position. A client can open a company’s tab to see which updates the AI considers important, with direct links to the original news or articles, showing that the AI has not fabricated information. Based on this data, the AI assistant also generates recommendations on whether it is worth investing in a particular company and explains why, using real sources.
– Does this mean the client is expected to independently review the information collected by the agent?
Andrew: – Not exactly. You can receive a summary of important news by email, Slack, Teams, or Telegram. The user can also set the frequency of notifications — every hour, once a day, or once a week. Unlike the previous version of the product, you no longer need to read lengthy texts; the summaries are just one to two paragraphs. If a user wants to dive deeper, they can open the dashboard and explore the detailed information there.
«The project’s ARR exceeds $200,000»
– What makes Altbridge different from other services offering investment advice?
Andrew: – Altbridge is consistent in its reasoning. To generate high-quality analytics, the AI agent analyzes 200–300 different sources.
– Who is the primary target user of Altbridge?
Nazym: – Our target clients are hedge funds and family offices (private companies that manage the finances and investments of wealthy individuals or families, note by Digital Business). These organizations employ analysts to cover specific tasks, but they often face a shortage of resources. Altbridge doesn’t replace employees; instead, it helps make each of them up to ten times more effective.
– How did potential clients react to your product?
Nazym: – The market needed time to adapt. Even large hedge funds in New York were initially skeptical about the solution. They had been working for decades within traditional models and were not ready for change. But now the situation is improving: many clients who, six months ago, were reluctant to engage have come back to discuss potential cooperation.
– How many clients do you have now?
Nazym: – We currently have more than five clients. These are primarily large companies, including one from Kazakhstan, as well as family offices from different countries.
– What is your business model?
Nazym: – Users purchase a subscription, with the fee determined by two factors: the number of companies included in the selection and the frequency of updates. Since these transactions involve high-value data, the cost ranges from $2,000 to $10,000 per month.
– So, have you reached profitability yet?
Nazym: – The project’s ARR is already over $200,000. With the new solution, however, we expect to scale much faster. Many clients are now waiting for the AI agent to become available to them.
«We plan to open a seed round in the United States this autumn»
– Earlier this spring, it was reported that you raised $750,000. How much has been invested in Altbridge overall since the beginning?
Nazym: – Venture growth requires significant investments that go beyond personal financial capacity. From the very beginning, we therefore focused on raising external funding. Initially, we didn’t plan to raise capital in Kazakhstan and instead intended to travel to the United States. Thanks to my work at the National Bank, I had built a broad professional network, so we had people to present Altbridge to. However, last April we met Murat Abdrakhmanov, who invested $200,000 in the project. At that time, the startup was essentially still at the idea stage—we only had a small prototype that analyzed podcasts.
Our second major investor was the Big Sky Capital fund, which also provided $200,000. As a result, our first tranche totaled $400,000.
A bit later, Altbridge also received investments from our network of friends, including GoatChat, former Chairman of the National Investment Corporation of the National Bank Eszhan Birtanov, and CEO of the nFactorial School Arman Suleimenov. Andrew Ang, an expert in factor investing (an approach where investors select assets based on specific factors and characteristics that have historically affected returns, note by Digital Business), also joined as both an advisor and investor. He is a former Professor of Finance at Columbia University and currently a Managing Director at BlackRock, the world’s largest investment company.
In total, following the pre-seed round, Altbridge has raised a little over $750,000 at a valuation of $15 million.
– How did you meet Andrew Ang?
Nazym: – I had known about Andrew since my time at the National Bank, but we first met when we tried to sell our prototype in the United States. We wanted to present it to BlackRock and explore possible collaboration.
Andrew Ang
Andrew explained that implementing such a solution at BlackRock would take too much time, but he liked what we were doing. He then shared two studies he was working on at the time and asked for our opinion. That’s how our correspondence began, and about nine months later Andrew wrote to us saying he would like to become our investor. So his decision was not spontaneous.
– You’ve raised funds from both Kazakh and foreign funds and business angels. How do investors differ depending on the region?
Nazym: – Our local investors have extensive experience, including international exposure, so in many ways they don’t differ much from U.S. fund representatives. However, I believe the industry in Kazakhstan is still maturing. The number of high-quality investors is growing, but it is still not as large as abroad.
– When are you planning your next round?
Nazym: – We plan to open a seed round in the United States this autumn. However, we take a careful approach to fundraising. Of course, the size of the investment matters, but even more important is the source of that funding. Any startup is, in a way, a story. If it is limited only to money, the project ends as soon as the financing stops. That’s why it is crucial to carefully select the right partners.
«We don’t aim to replace people — we aim to give them a powerful assistant»
– Who’s on the Altbridge team apart from you two?
Andrew: – Our team consists of five full-time employees, including us, and two part-time employees. In the beginning, we mainly hired ML engineers and mathematicians. Now we have started bringing in people with investment expertise. For example, not long ago, John Ospanov, an expert in value investing (an investment strategy focused on identifying undervalued securities through fundamental analysis, note by Digital Business) and former Chief Investment Officer at major Kazakh funds, joined our team.
– How will the product develop over the next year?
Nazym: – Our main priority now is to grow our client base, including in Kazakhstan. To achieve this, we are actively introducing Altbridge to as many analysts and portfolio managers as possible. We explain that there’s no need to be afraid of our product. Sometimes, when they see how fast and effective the AI agent is, they worry about losing their jobs. But we don’t aim to replace people — we aim to give them a powerful assistant.
Andrew: – But the current solution is only an intermediate stage. In the future, we want to develop Altbridge into a platform similar to Citadel, one of the world’s most powerful hedge funds and a multi-PM platform. This model brings together numerous independent teams of portfolio managers who trade within a single structure, using the same infrastructure, risk management system, and fund capital. The key difference is that, while at Citadel this work is done by people, at Altbridge these tasks will be carried out by an AI agent.