(C) Property Access

Domestic software companies that have long dominated the enterprise and institutional search solution market are currently weathering a harsh "growing pain" phase. As the global industrial landscape shifts toward Artificial Intelligence Transformation (AX), traditional search engine providers are struggling to balance the decline of legacy businesses with the aggressive R&D required to remain relevant in the generative AI era.

According to recent electronic disclosures, key players including Saltlux, Konan Technology, WiseNut, and Vaiv Company (formerly Daumsoft) reported continued operating losses or significant profit drops for the 2025 fiscal year. While the demand for AI is at an all-time high, these firms are caught in a pincer maneuver: their traditional search and chatbot revenues are shrinking, while infrastructure costs—specifically for GPU clusters and Large Language Model (LLM) development—are skyrocketing.

The Shift from Search to RAG and Specialized AI
Historically, these companies leveraged their expertise in managing structured and unstructured data to maintain a competitive edge in the chatbot market. When generative AI first emerged, they were among the first to address the "hallucination" problem by implementing Retrieval-Augmented Generation (RAG)—a technology that anchors AI responses in verified internal and external data.

However, the market is no longer satisfied with mere integration. To compete with global tech giants, domestic firms are forced to evolve from being "search providers" to "foundational AI architects," a transition that requires immense capital.

Company Performance Breakdown
Saltlux: Reported 2025 revenue of 41.64 billion KRW (down 9.4% YoY) with an operating loss of 8.04 billion KRW. The deficit widened by 21.3%. The company attributed this to delays in public/enterprise projects and preemptive investments in GPU servers.
WiseNut: Though it remained the only profitable firm among the four, its operating profit plummeted by 93.3% to a mere 110 million KRW. Excluding financial gains, the core business hovered near break-even due to IPO-related costs and R&D spending.
Konan Technology: Achieved a record revenue of 33.98 billion KRW (up 29.1%), but posted an operating loss of 9.87 billion KRW. Despite rapid growth, improving margins remains a critical hurdle.
 
Strategy for 2026: Efficiency and the Rise of "AI Appliances"
To pivot back to profitability, these companies are shifting their business models from labor-intensive "system integration" (SI) projects to standardized AI packages and hardware-integrated appliances.

WiseNut, for instance, plans to launch AI appliances powered by Neural Processing Units (NPUs). By moving away from expensive, generic GPU dependencies and toward specialized NPU hardware, they aim to lower the entry barrier for clients and improve their own cost structures. Their focus is on "AI Agents"—deployable, standardized models that can be integrated into a client's workflow with minimal additional development.

Similarly, Konan Technology is aiming for a turnaround this year. By leveraging their accumulated R&D, they have developed a high-efficiency architecture that significantly reduces the marginal cost of new project expansions.

Conclusion: The AX Breakthrough
The search solution industry stands at a crossroads. The transition from "finding information" to "generating intelligence" is a costly one, but industry experts believe the current deficits are a necessary investment for long-term structural improvement. As public sector AI projects resume and the demand for private, secure enterprise AI grows, these firms are betting that their "preemptive investments" will finally blossom into a sustainable AX-driven recovery.

Industry Note: The key to success in 2026 will be "Efficiency." Companies that can provide high-performance AI solutions without the staggering infrastructure costs of 2024-2025 will likely lead the next generation of the Korean software market.

[Copyright (c) Global Economic Times. All Rights Reserved.]