PER, EPS, PBR and BPS are calculated in-house from the last 4 quarters (2025Q3–2026Q2) · Dividend yield is based on ₩0 per share · Prices as of the 2026-10-02 close
SKAI, the only Korean company holding native graph database patent technology, is expanding into public-sector AI transformation projects and commercializing its ONTOVIA AI agent, but revenue has declined for three straight years and the company continues to post operating losses within a weak financial structure.
Founded in 2013 and listed on KOSDAQ in 2021, the only Korean firm holding native graph database patent technology
2025 revenue fell to about KRW 15.6 billion, the third straight annual decline, though the operating loss narrowed from 2024
The company turned briefly profitable in Q2 2025 before returning to quarterly operating losses from Q3 2025 onward
Expanding public-sector AI transformation contracts and selection as a supplier under the government AI voucher program
Ongoing dilution risk from convertible bond conversions increasing share count
SKAI (formerly Bitnine) was founded in 2013 and listed on KOSDAQ in 2021 as the only Korean company holding native graph database patent technology. Its core business model centers on licensing graph database software and providing related technical services.
A key technical differentiator is its hybrid architecture that combines relational and graph database structures. The company operates an R&D center in Silicon Valley to maintain global technical competitiveness.
Its product lineup consists of three offerings: AgensSQL, a PostgreSQL-based relational DBMS; AgensGraph, an enterprise-grade cross-model DBMS that processes both relational and graph data; and ONTOVIA, an AI agent combining graph RAG and vector RAG.
ONTOVIA is designed to minimize the hallucination problem common to generative AI by combining context-aware graph RAG with vector RAG.
The customer base skews heavily toward the public sector: of roughly 30 business contracts secured between January 2025 and March 2026, about 75%, or roughly 20, were AI and data projects with government and public institutions.
Confirmed major projects include a supply-chain early-warning system for the Ministry of Economy and Finance, a fraudulent-acquisition detection system for National Health Insurance Service collections, an upgrade to the National Police Agency's telecom-fraud response center system, and an AI platform build for Gyeonggi Provincial Government.
Overseas, the company also runs an AI-generated content (AIGC) solutions business under the 'SKAI Intelligence' brand.
| Quarter | Revenue | Operating profit | Op. margin |
|---|---|---|---|
| 2025Q2 | ₩6.9B | ₩2.6B | 38.5% |
| 2025Q3 | ₩2.5B | -₩2.7B | −106.0% |
| 2025Q4 | ₩4.2B | -₩1.3B | −32.1% |
| 2026Q1 | ₩3.2B | -₩1.7B | −51.1% |
| 2026Q2 | ₩2.7B | -₩2.4B | −88.4% |
| Year | Revenue | Operating profit | Net income (parent) | Op. margin | ROE | Debt-to-equity |
|---|---|---|---|---|---|---|
| 2022 | ₩19.2B | ₩1.6B | -₩400M | 8.5% | −1.3% | 107.3% |
| 2023 | ₩27.8B | -₩13.3B | -₩16.7B | −47.7% | −108.9% | 239.4% |
| 2024 | ₩19.8B | -₩12.1B | -₩24.8B | −60.8% | −239.3% | 169.6% |
| 2025 | ₩15.6B | -₩4.4B | -₩15.1B | −28.1% | −119.9% | 190.5% |
Consolidated basis (financial statements including subsidiaries) · Final figures from DART filings · Net income attributable to owners of the parent · Data as of 2026-09-05
According to the confirmed financials, SKAI's revenue rose from KRW 19.2 billion in 2022 to KRW 27.8 billion in 2023, then declined for three consecutive years to KRW 19.8 billion in 2024 and KRW 15.6 billion in 2025.
Operating profit swung from a KRW 1.6 billion gain in 2022 to losses of KRW 13.3 billion in 2023 and KRW 12.1 billion in 2024, before the loss narrowed to KRW 4.4 billion in 2025.
Net loss attributable to owners widened from KRW 0.4 billion in 2022 to KRW 16.7 billion in 2023 and KRW 24.8 billion in 2024, before easing to KRW 15.1 billion in 2025.
Operating cash flow moved from a positive KRW 2.9 billion in 2022 to outflows of KRW 9.9 billion in 2023 and KRW 11.8 billion in 2024, before the outflow eased to KRW 3.8 billion in 2025. The debt ratio jumped from 107.3% in 2022 to 239.4% in 2023, and has since stayed elevated at 169.6% in 2024 and 190.5% in 2025.
On a quarterly basis, the company briefly turned profitable in Q2 2025 with revenue of KRW 6.9 billion and operating profit of KRW 2.6 billion, but revenue collapsed to KRW 2.5 billion in Q3 2025, pushing the operating result back to a loss of KRW 2.7 billion.
Operating losses continued in the following quarters — KRW 1.3 billion in Q4 2025, KRW 1.7 billion in Q1 2026, and KRW 2.4 billion in Q2 2026 — bringing the combined net loss attributable to owners over the most recent four quarters (Q3 2025 through Q2 2026) to roughly KRW 25.2 billion.
Owners' equity increased from KRW 10.4 billion in 2024 to KRW 12.6 billion in 2025, suggesting external capital was raised even as losses continued.
Overall, annual losses appear to have bottomed out in 2024 and improved somewhat since, but quarterly results remain volatile and the brief return to profitability has not been sustained.
South Korea's public sector is expanding investment in data and AI infrastructure amid a government-led push for AI transformation (AX), with the AI voucher support program run by the Ministry of Science and ICT and the National IT Industry Promotion Agency (NIPA) serving as a key policy channel.
SKAI was selected as a supplier under the 2026 AI voucher program, giving it an opportunity to provide its solutions to government-subsidized small and mid-sized enterprises, medical institutions, and small business owners.
Graph database and ontology-based knowledge graph technology is drawing attention as a verification method to reduce hallucination in large language models, positioning it as a growth area alongside the spread of retrieval-augmented generation (RAG) technology.
However, competition is intensifying as large overseas vendors and domestic AI startups also enter the graph and vector database space.
SKAI differentiates itself as the only Korean company holding native graph database patent technology and is building a track record of proof-of-concept validations in secure, closed government network environments.
Because public-sector procurement is tied to budget cycles and project award timing, revenue recognition tends to cluster in specific quarters, contributing to an industry structure prone to significant quarterly variability.
In March 2026, SKAI was selected as a supplier under the AI voucher support program run by the Ministry of Science and ICT and NIPA, and is now working to identify public- and private-sector demand for ONTOVIA.
The company is also participating in a project commissioned by the National Information Society Agency (NIA) to establish a roadmap for the next-generation e-government standard framework, where it will conduct proof-of-concept validation of hybrid RAG and MCP technology suited to closed government network environments.
In June 2026, SKAI signed a partnership with TSID to jointly develop a zero-trust, security-enhanced AI database solution combining dynamic authentication technology with ONTOVIA, aiming to expand into markets such as finance and defense that require high-trust authentication.
In May, the company held its 'SKAI AI Strategy Seminar 2026' tour across major cities, presenting an AX transformation roadmap to roughly 200 public and educational institutions.
Overseas, under the 'SKAI Intelligence' brand, the company has introduced an NVIDIA Omniverse-based AI-generated content solution and, at GTC 2026, unveiled a retail-industry digital twin infrastructure strategy together with LVMH, signaling parallel efforts to expand globally.
That said, no specific revenue or profit guidance from the company has been confirmed, and sell-side reports offering an investment rating or target price on this stock are difficult to find.
Key items to watch going forward are how reliably public-sector contract wins convert into actual revenue and cash inflow, and whether the overseas AIGC business can establish itself as a separate revenue stream.
SKAI has posted net losses over each of the most recent four quarters, making earnings-based valuation metrics difficult to apply, and the stock currently trades on a multiple of net asset value rather than earnings.
Relative to net assets, the shares trade at a substantial premium, suggesting that growth expectations and the business-expansion narrative are weighing more heavily on the price than earnings fundamentals. The company pays no dividend, so comparisons on a dividend-yield basis carry limited meaning.
Given the elevated debt ratio and the company's reliance on external financing, including convertible bond conversions, to bolster capital, future changes in share count could affect how net asset value per share is calculated.
Looking at the multi-year pattern, the scale of operating losses eased somewhat after bottoming in 2023–2024, but the most recent quarters show losses widening again, so whether this improving trend continues remains to be confirmed.
PER, EPS, PBR and BPS are all calculated in-house (the same method as Naver and Toss) · Dividend yield = cash dividend per share from DART filings (supplemented by KRX) ÷ current price · As of 2026-09-05
SKAI is the only Korean company holding native graph database patent technology and developed the world's first hybrid graph DB architecture.
This technology combines graph RAG and vector RAG to reduce hallucination in generative AI and is building a proof-of-concept track record in secure, closed government network environments. This positions the company well amid the government's expanding AI transformation (AX) policy push.
Roughly 75% of the approximately 30 contracts secured between January 2025 and March 2026 were with public institutions, and the company was also selected as a supplier under the 2026 AI voucher program.
It has accumulated project experience across multiple government and public agencies including the Ministry of Economy and Finance, National Health Insurance Service, and the National Police Agency, which could support a more stable sales channel. Memoranda of understanding with academic and government-affiliated institutions continue to be signed.
Annual operating losses eased in 2025 after bottoming out in 2023 and 2024, and operating cash outflow narrowed in tandem.
At the same time, under the 'SKAI Intelligence' brand, the company has introduced an NVIDIA Omniverse-based AIGC solution and announced a collaboration with LVMH at GTC 2026, diversifying into overseas and private-sector markets.
The zero-trust security partnership with TSID can also be seen as an attempt to enter new high-trust markets such as finance and defense.
Annual revenue fell for three consecutive years, from KRW 27.8 billion in 2023 to KRW 19.8 billion in 2024 and KRW 15.6 billion in 2025. The company briefly turned profitable in Q2 2025 but returned to losses from Q3 onward, with operating losses continuing for four straight quarters since.
Because public-sector revenue tends to cluster around specific award timing, quarter-to-quarter variability is significant.
The company has posted net losses attributable to owners for four consecutive years since 2022, and its debt ratio, after spiking to 239.4% in 2023, has remained above 150% since.
Operating cash flow was also negative for three straight years from 2023 through 2025, indicating the business is not generating cash internally. The company continues to rely on external financing, including convertible bond conversions, to fund operations.
Sell-side reports offering an investment rating or target price on this stock are difficult to find, reflecting information asymmetry and limited market attention.
Additional common shares have continued to be listed following convertible bond conversions, and any remaining conversion volume could dilute existing shareholders further.
The debt ratio has remained above 150%, and operating cash flow has been negative for several consecutive years, making it difficult to cover operating funds solely from cash generated by the business.
As a result, the company has repeatedly relied on external financing such as convertible bond issuance and conversion, and further dilution of existing shareholders could recur if additional funding is needed.
A significant portion of revenue comes from government and public-institution projects, so revenue recognition tends to cluster around specific quarters tied to budget cycles and procurement timing. Changes in policy direction or budget allocation could directly affect the contract pipeline itself.
The graph DB, vector DB, and RAG-based AI agent market continues to see entry from large overseas vendors and domestic AI startups, which could intensify competition. Whether new solutions such as ONTOVIA can move beyond the pilot and proof-of-concept stage into stable, recurring revenue has not yet been confirmed.
Check the Q3 2026 earnings disclosure to see whether the recent trend of widening quarterly operating losses continues or reverses.
Track the demand-company matching results and actual revenue recognition timing from the AI voucher program selection made in March, to gauge how quickly public-sector contracts convert into revenue.
Monitor the commercialization progress of the zero-trust security AI database solution being jointly developed with TSID, and whether entry into new markets such as finance and defense materializes.
Watch for disclosures on further convertible bond conversions and additional common share listings to check for dilution from changes in share count.
Follow up on whether the overseas AIGC business under the 'SKAI Intelligence' brand, including the collaboration with LVMH, translates into actual contracts or revenue.
Built on its native graph database patent technology — the only such holding in Korea — SKAI is expanding contracts in the public-sector AI transformation market while attempting to diversify through its ONTOVIA AI agent and an overseas AIGC business.
However, the confirmed financials show revenue declining for three consecutive years since 2023, with operating losses and net losses attributable to owners persisting for four straight years and a debt ratio that has remained stubbornly above 150%.
The company briefly turned profitable in Q2 2025 but reverted to quarterly losses from Q3 onward, and the sizable cumulative net loss over the most recent four quarters raises questions about the durability of any earnings improvement.
The increase in share count from convertible bond conversions reflects a financial structure that has relied on external financing and carries the potential for further dilution.
With a growth narrative built around expanding public-sector contracts, the AI voucher program, and overseas AIGC partnerships coexisting alongside financial vulnerabilities such as declining revenue, persistent losses, and a high debt ratio, continued monitoring of quarterly results and the pace at which public projects convert into revenue will be important.
Report written 2026-09-05 · Data as of 2026-09-04
This content is AI analysis of market data and web search results, provided for information only. It is not a solicitation or recommendation to invest. Investment decisions and their consequences are the investor's own responsibility. Data may be delayed or contain errors.