Major Update for AI Meeting Minutes Tool 'LINE WORKS AiNote'

LINE WORKS Corporation (Headquarters: Shibuya-ku, Tokyo; President: Takeshi Shimaoka), which provides the business communication tool 'LINE WORKS' and various AI products, will update its AI minutes creation tool 'LINE WORKS AiNote' and begin offering it from Thursday, April 9th. As the sophistication of AI utilization in companies advances, such as task automation and decision support, there is a growing demand for meeting minutes and data to be utilized more quickly as 'information assets' that can be used for AI training data and organizational knowledge. Against this background, in this update, by combining new speaker identification technology (*1) with the high-precision 'speaker separation' function provided by 'LINE WORKS AiNote', we have added an 'Automatic Speaker Recognition' function that identifies the speakers during a meeting (up to 30 people per meeting *2) and automatically reflects the speaker's name in the meeting record. In addition, we have strengthened the 'AI Summary' function, making it possible to extract necessary tasks from the meeting record with a single tap. These functions will further reduce the time and effort required for checking and organizing during minutes creation, while also supporting the swift execution of actions after the meeting. Furthermore, alongside this update, we will begin offering an API for external integration that enables the retrieval and operation of data recorded in 'LINE WORKS AiNote' from your own systems or external services. *1: Speaker identification technology: Technology that identifies a speaker from the characteristics of their voice. *2: This means handling up to 30 candidate speakers per meeting as targets for identification through audio analysis. Since identification accuracy varies depending on conditions such as speech situations and audio quality, it does not guarantee that 30 speakers can be accurately identified simultaneously.

1. Addition of 'Automatic Speaker Recognition' function to identify and reflect speakers during meetings Using speaker identification technology, it automatically identifies 'who' made each statement in an internal meeting with multiple participants, and creates a meeting record with speaker information attached. Since it can recognize up to 30 people (*2) per meeting, it can be utilized even in meetings with a large number of people. This makes it possible to quickly create meeting records that clearly show 'who' said 'what' and 'when', which not only further reduces man-hours and improves efficiency in minutes creation, but also contributes to preventing cognitive discrepancies such as 'he said, she said'. [How to use] (Mobile version) 1. From the note, long-press the participant's name in the section of the target statement, and select the [Change Participant] menu. 2. Search for and specify the participant (speaker) from the address book. 3. From the next time onwards, the AI will identify the specified member and automatically apply it to the participant's name. *Administrator consent is required to use this function (first time only). *Sections where statements are made for a certain number of seconds or more are subject to speaker identification. In sections where speaker identification is possible, an 'AI icon' is displayed at the bottom right of the participant icon. *The target of automatic speaker recognition is internal members registered in the address book.