Have you noticed how capital still treats some regions as footnotes, even when the talent is already shipping products the rest of the world wants? That question sat with me while reading through the latest term sheet news out of Seoul. A group of investors just agreed to build a dedicated vehicle for later-stage companies across Central Asia, with AI native businesses sitting near the center of the brief. It is not a splashy mega-fund headline. It is quieter than that, and in my view that makes it more interesting.
Why This Silk Road Vehicle Matters Now
Qazaqstan Investment Corporation, IT Park Ventures, Mirae Asset Venture Investment and SparkLabs Group signed a term sheet to create a fund aimed at Series A and later companies. The working name is the SparkLabs Mirae Silk Road Fund I LP. Mirae Asset Venture Investment and SparkLabs would manage it. Kazakhstan’s national fund-of-funds and Uzbekistan’s technology-park investment arm are expected to come in as anchors, subject to the usual legal and corporate checks.
The timing was not accidental. The papers were signed during state visits by the presidents of Kazakhstan and Uzbekistan to South Korea, in the same week as the first Central Asia–Republic of Korea Summit. Commercial deals between South Korea and Kazakhstan alone were described as totaling nearly nineteen billion dollars. Against that backdrop, a venture vehicle looks small. It also looks like the piece that actually tries to move companies, not just commodities and infrastructure.
I’ve found that the most useful funds in emerging tech markets are the ones that admit a simple truth: early proof is no longer the scarce thing. Distribution is. The mandate here is built around that idea. The team wants companies that already have a working model and are ready to leave home markets, not garage experiments still hunting for product-market fit.
Who Is Putting The Structure Together
SparkLabs has backed more than six hundred startups across six continents since 2013. That kind of volume does not guarantee taste, but it does mean the firm has practiced the unglamorous work of helping founders open doors in new cities. Mirae Asset Venture Investment sits inside a financial group that manages hundreds of billions of dollars across eighteen markets. Put those two together and you get a mix of operator-style acceleration and institutional balance-sheet comfort.
On the limited-partner side, Qazaqstan Investment Corporation functions as Kazakhstan’s national fund of funds under Baiterek National Managing Holding. It already sits in nineteen funds with combined capitalization around 2.8 billion dollars, usually alongside private and international partners. IT Park Ventures is the venture arm of Uzbekistan’s state technology park. It backs companies from Uzbekistan and neighboring markets while trying to plug them into foreign capital.
Separate figures circulating around the agreement put initial capital at two million dollars, with room to grow toward ten million. That is modest by Silicon Valley standards. For a first regional vehicle with public-sector anchors and a cross-border expansion thesis, it is a starting line, not a ceiling. IT Park Ventures has talked about coming in as an anchor in stages, which is the sort of cautious phrasing you hear when lawyers still have work to do.
Higgsfield becoming Kazakhstan’s first unicorn was a turning point. It showed that a world class AI company can be built in Central Asia.
– Aslan Sultanov, co-founder and general partner
That quote is doing a lot of work. It is not just celebration. It is a pitch to limited partners who still need a proof point before they treat the region as more than a frontier experiment.
What The Mandate Actually Looks Like
The fund is sector agnostic on paper. In practice, the language keeps circling back to companies where artificial intelligence sits inside the product, not as a slide-deck accessory. Think tools, platforms, and workflows that stop making sense if you strip the models out. That is what people mean by AI native in this context.
Stage is clearer than sector. Series A and later. Proven model. Ready to expand. Those three filters cut out a lot of noise. They also explain why the check size can stay relatively contained at first. Later-stage rounds in the region do not always need Bay Area tickets. They need patient capital plus a map.
- Target geography: startups across Central Asia, not a single-country bet
- Stage: Series A and companies already past early product risk
- Focus: AI-native businesses with room to sell outside home markets
- Support: introductions and operating help into South Korea, the United States, and the MENA region
- Structure: managers from SparkLabs and Mirae Asset, anchors from Kazakhstan and Uzbekistan
Portfolio companies would tap the networks of both managers. That is the real product. Money is the ticket. The aftercare is the reason a founder might take a smaller check from this vehicle instead of waiting for a larger, more distant fund that will never pick up the phone after the wiring is done.
The Higgsfield Effect And A Young Population
Higgsfield crossed a billion-dollar valuation in 2025 and became Kazakhstan’s first technology unicorn. The company builds AI video generation tools. Its founders include Yerzat Dulat and Alex Mashrabov, who previously led generative AI work at a major consumer platform. Whether you love the product category or not, the valuation changed the conversation. Suddenly the region had a reference company that outsiders could pronounce.
Sultanov also pointed to demographics. More than half of Central Asia’s population is under thirty. Governments are putting money into AI programs. International venture dollars still lag what you see in more crowded tech markets. That gap is either a warning or an opening, depending on your temperament. I lean toward the second reading, with a caveat: youth plus policy speeches do not automatically produce durable companies. Execution still has to show up.
Perhaps the most interesting aspect is how quickly a single unicorn becomes a recruiting tool. Founders in neighboring countries now have a story they can tell employees and customers. Investors have a chart they can drop into a memo. That is how regional narratives get built, for better and worse.
Seoul, Summits, And The Politics Of Capital
State visits produce photo ops. They also produce term sheets that would have taken another year in ordinary calendar time. South Korea has spent years deepening commercial ties with Central Asia. Energy, logistics, manufacturing, and now software sit on the same table. A venture fund is a small line item next to multi-billion industrial packages, yet it is one of the few instruments that can follow a founder across borders.
There is a practical reason Korean capital cares. Korean corporates and funds want deal flow that is not already picked over in Seoul, Silicon Valley, or Singapore. Central Asian teams often price more reasonably, move faster on operations, and still need help with brand, compliance, and enterprise sales in richer markets. That is a classic arbitrage, if you can underwrite governance and currency risk without getting romantic about it.
In my experience, summit-week funds either become real platforms or fade into memorandum status. The difference is usually whether someone opens an office and starts taking meetings in Tashkent and Almaty on ordinary Tuesdays, not only when cameras are present. Plans around this agreement include a SparkLabs presence at IT Park in Tashkent. That detail matters more than the branding.
Kazakhstan’s Parallel Push Into Digital Assets
The AI story does not sit alone. Kazakhstan has been building other pieces of a digital economy at the same time. A technology community project associated with a well-known former exchange executive and investor reopened in the country, pitched as a base for founders and mobile operators. A major messaging platform opened its first regional office there and launched an AI lab. Those moves do not create a cluster by themselves. They do signal that international operators are willing to plant a flag.
Government policy has also reached into digital assets. Authorities outlined a national analytics center on the central bank’s supervisory technology stack. The idea is to watch fiat and crypto flows, wallets, and customer data, then give banks, law enforcement, and licensed providers better verification tools. That is a compliance build-out, not a free-for-all. Founders who want to sell into regulated finance will care about that architecture even if they never touch a token.
Earlier in the year, officials approved a framework for large-scale miners. Qualifying operators can get structured access to power if they meet infrastructure rules and contribute part of mined assets toward a state-backed reserve. Seized digital assets had already been discussed as one possible source for that reserve, alongside more traditional holdings. You can argue about the politics of that design. You cannot argue that the state is ignoring the sector.
Uzbekistan’s Own Infrastructure Bet
Uzbekistan is the second public-sector anchor in the new vehicle, and it has been writing its own rulebook. A regulated mining zone in Karakalpakstan offers approved firms access to several power sources and tax incentives running through 2035. Operators can use grid power, renewables, and hydrogen. Sales proceeds from mined assets are expected to stay inside the domestic banking system. That last point is easy to skip. It is actually the heart of the policy: keep the cash visible.
The Beshkala Mining Valley framework covers licensed operations across the region and allows sales through foreign platforms under local rules. There is even a side note about using excess heat from mining halls for greenhouse farming. It sounds quirky until you remember how often industrial policy tries to stack two problems into one site.
IT Park Ventures is using the Silk Road vehicle as a bridge. The Tashkent office plan is meant to help local companies meet foreign managers without flying halfway around the world for every partner meeting. That sounds basic. It is also how most regional ecosystems actually get wired.
How This Fits A Broader Shift In Venture Taste
Later-stage concentration is not a Central Asian quirk. Across digital-asset and adjacent tech investing, capital has been clustering in companies that already have revenue stories. One research snapshot put second-quarter crypto and blockchain fundraising at 5.683 billion dollars across 384 deals, with later-stage firms taking 78 percent of the dollars. The first half of the year reached just over ten billion dollars across 744 transactions. U.S.-headquartered companies still absorbed the majority of capital in the second quarter.
That last figure is the uncomfortable one for any regional fund. If three quarters of the money still lands in one country, a Silk Road vehicle has to offer something those large funds will not: proximity, political cover, and a path into Korea and the Middle East that does not require a founder to relocate on day one.
Other managers have been blending AI with neighboring themes such as robotics, energy, and on-chain infrastructure. The pattern is familiar. Once models became a default layer, pure-play labels started to blur. A Central Asian company that uses models to run logistics, media, or industrial software can look like an AI deal, a regional deal, and an export deal at the same time. Funds like this one are built for that overlap.
| Piece | What We Know | Why It Matters |
| Managers | SparkLabs and Mirae Asset Venture Investment | Operator network plus institutional reach |
| Anchors | QIC and IT Park Ventures | Local political and ecosystem access |
| Stage | Series A and later | Less product risk, more expansion risk |
| Theme | AI native, sector flexible | Models as core product, not a feature |
| Initial size | About $2 million, path to $10 million | A first close, not a final statement |
What Founders Should Actually Expect
If you are building in the region, the useful question is not “Is this fund big enough?” It is “Can these people help me sell in Seoul, Dubai, or a U.S. enterprise account without wasting a year?” Expansion support is easy to promise and hard to staff. The firms involved have done it before in other markets. That is necessary. It is not sufficient.
Founders should also expect slower legal work than the press cycle. Anchor commitments that depend on approvals can slip. Currency rules, shareholder agreements, and cross-border holding structures in this part of the world are not copy-paste from Delaware. A good manager will say that out loud. A marketing deck will not.
- Show a model that already works in at least one home market.
- Explain why the next geography is South Korea, the United States, or MENA rather than “everywhere.”
- Be ready to discuss data, compute, and talent costs with adult numbers.
- Treat governance as part of the product if you sell to banks, telcos, or the state.
- Ask who, by name, will make the first ten introductions after the money lands.
That last item sounds blunt because it should. Networks are not abstract. They are calendars.
Risks That Do Not Fit On A Launch Slide
Public-sector anchors can open doors and they can slow decisions. Both things are true. A national fund of funds brings legitimacy. It also brings process. International limited partners will watch that tension closely. So will founders who have lived through committees that move at the speed of a stamp.
There is concentration risk too. If too much of the story hangs on one unicorn and a handful of AI labs, the narrative gets fragile. Regions need a second and third breakout, preferably in different categories, before outsiders stop treating success as an exception.
Currency, talent flight, and compute access remain ordinary problems. Models are hungry. Power policy that works for miners may not automatically work for inference clusters in cities. I would want to see the fund’s operating partners get specific about cloud, chips, and bilingual product design before calling the thesis complete.
And then there is the size. Two million dollars does not reshape a market. Ten million still does not, not by itself. The vehicle works if it becomes a trusted first institutional check that later global funds are willing to follow. If it tries to be the whole stack, it will look thin.
A Personal Read On The Opportunity
I’ve sat through enough emerging-market launches to know the pattern. Big flags. Modest first close. A promise that this time the region is “next.” Sometimes it is. Sometimes the office quietly turns into a listening post and the second fund never appears. The difference is usually boring: weekly pipeline meetings, local partners who can say no, and a willingness to pass on politically convenient deals that will not export.
What I like here is the honesty of the stage filter. They are not pretending every student hackathon is a Series A. What I will watch is whether AI native becomes a useful screen or a slogan that lets anything with a chatbot through the door. The region does not need more wrapper products. It needs companies that can survive a sales cycle in a foreign language.
Is Central Asia about to become a default stop on every global AI tour? No. That would be a lazy forecast. Can a well-run, politically plugged-in vehicle pull a handful of companies into Korea, the Gulf, and U.S. customer lists faster than those companies could manage alone? That is a fairer bet, and it is the only bet this fund needs to win in its first life.
What To Watch After The Term Sheet
Term sheets are not funds. The next markers are ordinary and public enough to track. Did the anchors complete their approvals? Did the Tashkent office open with actual staff? Did the first two investments look like expansion stories or hometown favors? Did Korean and Middle Eastern customers show up, or only Korean and Middle Eastern conferences?
Watch also whether the vehicle stays disciplined on stage. The temptation in a thin market is to drift earlier because that is where the volume sits. Drift is how theses die. If the team keeps writing later-stage checks into companies that can show revenue and a border plan, the modest first close will look smart in hindsight.
There is a human piece as well. Young engineers in Almaty, Tashkent, Bishkek, and beyond have been shipping quietly for years. A unicorn gave them a headline. A fund gives them a counterparty. Neither replaces the grind of building something a stranger will pay for. But counterparties matter. They always have.
Capital follows proof, then it follows a path. This vehicle is trying to supply the path.
That is the cleanest way I can put it. The proof, at least in one category, already exists. The path runs through Seoul as much as it runs through Central Asia. If the managers treat that route as a job rather than a slogan, this small fund could punch above its first-close weight. If they do not, it will join the long list of summit souvenirs. I know which version I would rather read about a year from now.
A Longer View For Operators And Allocators
Allocators scanning emerging Europe and Asia already have more regional vehicles than they can diligence in a quarter. Another logo does not help them. A vehicle that can show follow-on interest from Korean strategics, Gulf family offices, or U.S. growth funds after the first two years would help them. That should be the internal scoreboard, even if it never appears in a press note.
Operators should treat the fund as one instrument among several. Government parks, corporate development teams, and customer-led expansion often move a company farther than a single check. The useful combination is a local anchor who understands procurement plus a foreign manager who can translate the product into another market’s buying language. On paper, this pairing tries to be that combination.
Will every AI-native firm in the region fit? Of course not. Some products are culturally specific. Some regulation will not travel. Some teams will decide that staying domestic is the higher-return path. That is fine. A fund like this one should be allowed to be picky. Pickiness is the feature.
So here we are. A term sheet in Seoul. Two public institutions from the steppe and the Fergana basin. A Korean financial group and a multi-continent accelerator. A small pile of capital aimed at companies that already work. It is not the loudest story in markets this week. It might still be one of the more revealing ones, because it shows where patient money thinks the next exportable software will come from when the obvious zip codes get too expensive.
I’ll keep an eye on the first investments. Not the ceremony. The cap tables. That is where you learn whether a Silk Road label was geography, marketing, or an actual route.
