Yeongseon Park

ATAD Corp. does not sell servers. It sells the feeling of finally seeing your own infrastructure. Park, the founder and CEO, runs the company out of Seoul with a U.S. arm, ATAD Technologies Inc., and a flagship product named ODiiN, after the Norse god who traded an eye for the ability to see everything. The mythology is not an accident. The whole pitch is visibility.

ODiiN is described as an AI multi-cloud operating platform: one panel that operates AWS, Azure, and GCP together, with a real-time guide aimed squarely at people who are not cloud engineers and never wanted to be. The promise is optimization and automation that maximizes stability while minimizing cost. ATAD's own number is the loud one: up to 80% savings, with intelligent resource forecasting doing the quiet work underneath.

That number is the kind of claim that either ends a conversation or starts one. Park clearly prefers the second. ATAD's published metrics read like a wall of evidence stapled to the door: 380+ clients, a 4.8 out of 5.0 satisfaction rating, more than 15 million data transactions, over 100 integrations, and a paid conversion rate it lists at 90%. Take the figures as the company states them. The interesting thing is the confidence to publish them at all.

The second product tells you where Park's head is pointed next. Heiimdahl, named after the gatekeeper of the gods, is a multi-LLM monitoring platform pitched as a secure AI gateway, one that filters threats and hallucinations in real time and routes each query to whichever model is best for it. In a year when every company bolted a chatbot onto something important and prayed, a gatekeeper that watches the models is a tidy read on the moment.

Under both products sits the part Park keeps coming back to: data that belongs to the customer and stays that way. ATAD describes decentralized customer-data protection built on a stack of AI, multi-factor authentication, and zero-knowledge proofs, the cryptographic trick that lets you prove something is true without revealing the thing itself. It is a Web3 instinct wearing an enterprise suit. The keywords trail behind the company like a comet: blockchain, multi-cloud ecosystem, decentralized data, ZKP, automation for non-professionals.

The numbers ATAD wants on the record

Then there is the roadmap, which Park does not whisper. ATAD has raised roughly $3.36 million across multiple venture rounds, including a Series A that brought in about $2.52 million in May 2025. The stated destinations: a Nasdaq listing by 2027 and decacorn status by 2028. Most founders keep that arithmetic in the deck and out of the press. ATAD puts it on the table. You can call it brash. You can also call it accountability with a date attached.

What makes the story worth following is the framing. Plenty of companies promise to tame the cloud for experts. Park aimed lower and harder: the non-expert who still, somehow, has five clouds to run on a Tuesday. That is a bigger audience than the priesthood of certified architects, and a more honest description of how most businesses actually live. The expertise gap is the market. ODiiN is the bridge.

He does not run it alone. ATAD's masthead is a tight trio out of Seoul: Park as CEO, Youngwook Jeon as CTO, and Sujin Jung as CPO. The company calls itself a builder of a sustainable value-chain, and the naming habit, gods and gateways and a decentralized VPN project called Kraton, suggests a founder who treats ambition as a design language. You name a product Heiimdahl when you intend it to guard something that matters.

The honest caveat belongs here too. ATAD is early. The metrics are self-reported, the Nasdaq date is a target and not a ticker, and the cloud-management field is crowded with giants who would happily absorb the idea. None of that is hidden, and Park does not seem interested in hiding it. The bet is simple and a little stubborn: complexity is not a law of nature, it is a product problem, and product problems can be fixed by someone willing to stare at the invoice long enough to get angry about it.

So here is where to catch him, mid-stride: a founder in Seoul, shipping an AI autopilot for the multi-cloud mess, naming his tools after the watchers and gatekeepers of old myth, and telling anyone who asks exactly which year he plans to ring the bell. Whether 2027 arrives on schedule is the open question. The clarity of the ask is the answer he is already giving.