Due Diligence
I have spent eight years building a no-code AI platform for large scale imagery from scratch. No external funding. No PR agency. No deck full of promises about a product that doesn't exist yet.
What I have is a working product, paying clients, and reference deployments spanning aerial imagery and microscopic image analysis. The kind of thing that takes years to get right.
And more than once, I’ve watched evaluation panels side with competitors who were willing to lie far more.
Pattern 1: Funding as a Proxy for Credibility
Here is how it usually goes.
A pitch event evaluation or government grant selection happens. One team has a functioning product with documented deployments. Another team has a polished pitch, mentions the buzzwords, and has received venture funding — which, in many evaluators' minds, is treated as a PROXY for credibility.
The funded team wins, or at minimum gets treated as the more serious contender.
What the evaluators don't know — because nobody checked — is that the funded team has no production-ready product. In some cases, they had previously approached the builder to discuss OEM arrangements. They wanted to resell the actual working technology under their own brand. That conversation went nowhere. Months later, they showed up as a competitor in the same evaluation.
No one on the evaluation panel knew any of this. No one asked to see a working demo on real data. No one requested deployment references beyond what was written in the proposal. The pitch was taken at face value.
This is not one story. This is a pattern.
Pattern 2: Circular Revenue Built on Public Money
A professor instructs his students to establish a company.
The professor then applies for research funding through a consortium that includes the student’s company. From time to time, the professor receives consulting fees from that company, while also subcontracting work to it, allowing the student’s company to generate revenue.
Once those sales are reported to the tax authorities, the student’s company can present itself as a promising startup with strong revenue. On that basis, it may attract investment and receive additional government support.
The student, in turn, uses investment capital and public subsidies to purchase the professor’s research outputs, reports, or related services.
In this structure, revenue from the real-world market is no longer necessary. By circulating public funds, government grants, investment money, and subcontracting payments among themselves, they can create a self-sustaining ecosystem funded by taxpayers. They effectively become parasites on public money, supporting one another through commissioned work and artificial revenue.
What they need is not customers, but more students. Each additional student-founded company becomes another serpent attached to the head of Medusa, expanding the network and increasing the amount of money that can be captured.
Even if they cannot earn money from genuine markets or real customers, they can still obtain funding. By using public grants, subsidies, and investment capital, they continue to conceal one falsehood with another. They then use that same money to purchase reports, research outputs, and services that help cover up the weaknesses in their business, allowing the scheme to continue under the appearance of legitimate research, entrepreneurship, and innovation.
Why Due Diligence Becomes Superficial
To be precise, due diligence does take place in the Korean startup market, at least in a weak and superficial form. Venture capitalists can usually check the sales tax invoices of the companies they invest in. They know where the revenue is coming from. They know whether the company is losing money. They know whether profitability is realistically going to improve in the short term.
The issue is that they do not conduct due diligence to find out whether the company is genuinely sound. They conduct it to protect their own exit.
The so-called VCs in Korea are mostly people who parade around with elite university credentials in Korea’s credential-obsessed society. But very few of them have actually worked as developers, built real companies, or operated businesses as founders. Most of them do not have meaningful domain knowledge in the industries they invest in. They do not understand the technology. They do not understand the market. They do not understand the customers.
What they do understand is whether a company can keep selling the lie long enough.
Because their own money is on the line, they still check whether the company is good at bluffing. They want to know whether the startup can package itself well enough to raise the next round, reach an IPO, and eventually dump its shares onto retail investors.
Their goal is not to build a good company. Their goal is to pass the company to the next investor, or to dump the shares on individual investors after listing. So their “due diligence” is really an assessment of the founder’s ability to keep lying until the next exit window opens.
In other words, they are not asking, “Is this a real business?”
They are asking, “Can these frauds keep the story alive until we can get out?”
That is why the ability to scam, manipulate numbers, dress up weak revenue, and tell a convincing growth story often matters more than real technology, real profitability, or real market demand. The company does not need to be good. It only needs to look investable until someone else buys the equity.
And they do not care whether retail investors lose money in the Korean market. Once the shares are passed down the chain, the final losses are someone else’s problem. If ordinary people get burned, that is irrelevant to them.
The structure is even worse because many startup funds in Korea are built with public money. Even when the startup investment fails, these funds are often backed by government capital. So the VCs get to play investor with taxpayer-supported money, collect management fees, pretend to be sophisticated financiers, and then walk away when the whole thing collapses.
That is why due diligence in this market is not real due diligence.
It is a fraud-screening process run by people who are not trying to avoid fraud, but trying to identify frauds that are competent enough to survive until the next round, the IPO, or the final dump onto retail investors.
What It Costs
The obvious cost is paid by the organizations that make bad procurement decisions. They buy solutions that underdeliver, then spend additional budget trying to repair the damage.
But the deeper cost is paid by honest builders and ordinary people.
In this kind of market, people who do not lie lose. Startups that are actually trying to build something real lose. Founders who refuse to exaggerate their technology, inflate their numbers, or participate in collective fraud with investors and the media lose. They are rejected not because their products are worse, but because they are not willing to perform the kind of deception that certain investors prefer.
A startup can lose in grant evaluations simply because it did not tell a big enough lie. It can be ignored by investors because it refused to dress up an unfinished product as a world-changing platform. Meanwhile, companies that are better at fundraising, media manipulation, and narrative inflation are rewarded as if they were the future.
The victims are not only founders.
Ordinary retail investors also pay the price. They read media coverage describing some AI or space startup as promising after its listing. They believe the story. They invest their savings. Then the company issues new shares, carries out capital reductions, gets delisted, or simply keeps losing money until the stock price collapses. By the time the truth becomes visible, the people who promoted the story have already benefited, and the losses have been pushed onto the public.
This is not just a financial problem. It destroys trust.
When lies are rewarded and honest work is punished, the market stops being a place where better products win. It becomes a place where better fraud wins. Over time, people stop believing in companies, investors, media, and even public institutions. The entire system begins to teach one message: corruption pays, deception is strategy, and honesty is stupidity.
That is the real damage.
A society that repeatedly rewards fraud eventually produces more fraud. It turns lying into a professional skill. It turns manipulation into a business model. It convinces people that moral decay is not a cost, but a path to profit.
Why Keep Mentioning Investors
The reason I keep mentioning investors is simple: in Korea, if a startup cannot raise investment, it is often locked out of startup grants, subsidies, government support programs, and government-backed loans.
Many export support programs, SME business development support programs, R&D funds, and public startup support schemes are either directly reviewed by venture capitalists or heavily influenced by whether the company has already received venture investment. In practice, this creates a tilted playing field. Companies that have been selected by investors are treated as more legitimate, while companies that have not raised funding are often treated as less credible, regardless of whether their products are actually better.
I raised this issue with Korea’s Ministry of SMEs and Startups. The answer I received was that the ministry itself does not have enough technical expertise, so it delegates evaluation to venture capitalists who supposedly have that expertise.
That response exposes the problem.
If the evaluators are investors whose incentives are already distorted, then the public support system inherits the same distortion. Government grants, subsidies, loans, R&D funding, and export support begin to reflect the preferences of venture capital firms rather than the actual quality of the technology, the product, or the business.
What I Lost — and What Others Will Lose
Over the past eight years, I have been fighting to make our go-to-market strategy work.
I still regret trying to bootstrap my way into the public sector and remote sensing market. But I have not given up on the business yet, because there are still things I can try, and there are still customers who need what we have built.
During those years, I watched many frauds appear and disappear.
I have not seen many of them build companies that eventually generated real operating profit. Meanwhile, I am still fighting, but our company’s cash flow is improving more than I expected.
What I seem to have lost over these eight years is trust: trust in this society, and probably trust in people as well.
There are many other things I have lost and given up along the way, but I do not want to talk about all of them.
Maybe this kind of society exists no matter which country I live in. But at least in Korea, I believe the market has been severely contaminated by excessive government support and dependence on venture capital.
Ten years ago, biotechnology companies were aggressively listed in Korea. Most of them eventually failed to develop new drugs, failed in clinical trials, or collapsed. Investors probably knew very well that it would not be easy for startups with insufficient technology, insufficient talent, insufficient funding, and insufficient experience to develop actual new drugs.
But the investors did not take the real losses.
Retail investors did.
Some lost their life savings. Some lost everything. Some even took their own lives.
Now another theme is being promoted in Korea: aerospace.
But we already know that even companies like Maxar and Planet, which listed in the United States years ago, have struggled with persistent losses. Yet in Korea, space startups that survive on government support are now rushing toward public listings.
The next biotech-style collapse may come from the space sector.
And when it happens, no one will take responsibility.
Due Diligence for Our Product
I am still trying to protect the integrity of my own product: its due diligence, its debugging process, its functionality, and the quality of its inference results.
For customers, this means one simple thing: we do not ask you to believe a pitch. We ask you to test the product on your own data, under your own constraints, and judge the result directly.
What I have now is a few computers, a stack of unmerged pull requests, and a backlog of issues waiting to be solved.
What I want, in the end, is simple: when my life is over, I hope there will have been more happiness than pain.
So even inside this painful business, I keep doing the work I owe my customers, and I keep trying the things I still need to try for my own life and my business.
Compared to the past, I no longer place the business above my health or my life. I am trying to find a better balance. But harsh conditions and constant stress continue to test me.
Still, just as I have endured countless trials so far, I intend to endure today’s storm as well.
Contemnit procellas.
