On November 18, 2026, FIXER launched 'GaiXerScribe,' a meeting minute creation application that runs on its on-premise generative AI appliance, 'Sovereign GaiXer.' The product is positioned to support minute-taking operations in environments handling confidential information, such as corporations, local governments, and medical institutions.

A key feature of GaiXerScribe is that it leverages NVIDIA's speaker diarization model, 'Nemotron 3 Diarization,' implemented on Sovereign GaiXer via API, to transcribe conversations while identifying 'who spoke when.' Sovereign GaiXer is an on-premise generative AI appliance that can complete AI processing without transmitting internal data externally, and this GaiXerScribe also features a configuration designed for deployment in highly confidential environments.

GaiXerScribe supports microphone recording, importing audio and video files, and text input, returning transcripts with speaker labels and timestamps. Users can modify or merge speaker names and correct spoken content while reviewing the original statements and audio.

Meeting minute output is based on eight categories: 'Meeting Information,' 'Summary,' 'Decisions,' 'To-Do,' 'Discussion Points,' 'Q&A,' 'Pending Items,' and 'Next Meeting,' and template settings can be customized according to user needs. Furthermore, it includes a mechanism to standardize the notation of technical terms and proper nouns by utilizing organizational or personal glossaries. It allows users to refer to the original statements that form the basis of each minute item and supports checking for discrepancies in numbers and dates, making it designed to be easily organized into 'actionable records' rather than just simple transcripts.

It also supports saving edit history and restoring previous versions, and meeting minutes can be exported in Word, PDF, TXT, and Markdown formats. Decisions and To-Do items can be exported as CSV files including assignees and deadlines, and it provides settings for internal sharing, management of retention periods, and operational audit logs.