Dinotisia has laid the groundwork for entering the public-sector market. Dinotisia said on Jan. 28 its vector database solution, Seahorse, received Grade 1 GS certification, the highest level of software quality certification, from the Telecommunications Technology Association (TTA). Seahorse goes beyond simple data storage to cover an AgentOps environment that supports the full lifecycle of building and operating AI agents.

GS certification is a national programme that tests software in real operating environments for functional suitability, performance efficiency, usability, reliability and security based on ISO/IEC international standards. With the Grade 1 certification, Seahorse meets the requirements to be designated for preferred purchasing by public institutions. The company said it has secured technical and institutional requirements for entering the public market, including third-party unit price contracts with the Public Procurement Service.

Seahorse is designed on Kubernetes to minimise constraints from infrastructure environments. It allows flexible model deployment and scaling. Companies can start with the cloud in the early stage of a service and switch to on-premises as data grows, and they can also pursue a hybrid strategy that runs both environments at the same time.

Dinotisia applied MCP (Model Context Protocol), an AI industry standard, to Seahorse. It implemented an open AI agent architecture that minimises dependence on a specific large language model (LLM) or data structure. It can build an AI agent environment that is easy to expand by connecting various LLMs and data sources through a standard interface.

Based on the certification, Dinotisia will seek to register the Seahorse AgentOps solution on the Public Procurement Service's Digital Service Mall in the first half of this year. It aims to support public institutions in adopting AI services through simplified procedures without complex contracting, and to use Seahorse as core data and agent infrastructure underpinning AI transformation in the public sector.

Dinotisia CEO Moo-kyung Jung (정무경) said, "Model performance is important to the success of an AI business, but it is often determined by operational reliability." He said, "With this Grade 1 GS certification, Seahorse has been verified to be usable even in business environments that demand high reliability." He added, "This year, Dinotisia plans to provide reliable AgentOps across the entire process for on-premises and cloud and contribute to companies' AI transformation (AX)."