Verified, then logged.
Sun 22 Nov 2026 · 00:17 KST / 15:17 UTC한국어 | ENSearch /

02 / IT · Analysis

STORM Agents shares work knowledge across AI agents

The service connects chat requests to business systems and requires staff approval for actions that affect external parties.

AI-generated editorial illustration for this article

Sionic AI, a Korean AI company, unveiled STORM Agents for enterprise users to share work knowledge between agents, according to MoneyToday, a Korean news publication. The service connects requests in chat to actions in business systems.

According to MoneyToday, the shared knowledge includes methods learned while carrying out tasks. This allows knowledge gained by one agent to inform another agent's work.

MoneyToday reported that agents return their results through chat. The service pairs this workflow with access controls for each agent and staff approval for actions that affect external parties.

Work results return through chat

According to MoneyToday, users can submit requests through Slack, internal messaging services and groupware, or shared workplace software. Agents then operate business systems to carry out those requests.

MoneyToday reported that agents deliver their output to the chat window. Chat therefore serves as both the starting point for a request and the place where users receive the result.

Approval for actions affecting external parties

Access permissions. According to MoneyToday, STORM Agents manages access rights for each agent. Knowledge sharing is paired with controls at the individual agent level.

Staff approval. MoneyToday reported that actions affecting external parties require approval from the responsible staff member. Sending external emails falls within this process. The workflow therefore connects chat requests to execution while retaining a human approval step for these actions.