Meet AI+
AI meeting minutes, run in a fully closed environment. Recording, speaker recognition, and summarization all happen within a closed environment, reducing the effort of creating meeting minutes.
What is Meet AI+
Meet AI+ is an AI meeting minutes product that runs in a fully closed environment.
From recording file upload through speaker recognition to AI summarization, everything takes place within a closed environment. It meets requirements where conversations cannot be shared externally. Built on our Generative AI platform, with an approved LLM deployed in-house.
Do any of these post-meeting burdens sound familiar?
- Creating meeting minutes takes time. Replaying recordings, sorting out who said what, and pulling together the key points takes more than an hour every time.
- Conversations can't leave the building. Information governance requirements mean you can't entrust meeting data to cloud-based transcription services.
- It's hard to tell who said what. In meetings with multiple participants and multiple languages, just distinguishing speakers to keep an accurate record takes effort.
From recording to meeting minutes in 3 steps.
Just hand over the meeting audio, and a formatted set of minutes comes out the other end.
- 01
Import the recording
Upload an audio or video file. Also supports Microsoft Teams and calendar integration, so you can pull in meeting recordings directly.
- 02
Speaker recognition and transcription
Transcribes Japanese, English, and Vietnamese speech while distinguishing who is speaking. Names can be assigned to each speaker.
- 03
AI summary and minutes output
Automatically summarizes the key points even from long meetings, and outputs minutes in a format matched to the meeting type.
* Transcription and summarization accuracy vary depending on the usage environment (based on our own findings).
Key features
Fully closed operation
Runs in a closed environment with an approved LLM deployed in-house. Can be deployed on-premises or within your own environment, operating to match your information governance requirements.
Speaker recognition (JA/EN/VI)
Distinguishes who is speaking and recognizes Japanese, English, and Vietnamese. Assigning names to speakers and search functionality improve the accuracy and usability of the minutes.
AI summarization
Automatically summarizes the key points even from long meetings (* accuracy varies depending on the usage environment).
Meeting-type templates
Minutes format can be customized to match the type of meeting or department, such as regular meetings, sales negotiations, or on-site meetings.
Significantly cut the time spent creating meeting minutes.
We also address the closed-network and information governance requirements of IT departments, starting from our approach to security.
See our approach to securityDeployment in as little as 2–4 weeks.
Deployment process
Requirements discovery
We ask about your information governance requirements, target meetings, and expected number of users, then propose an operating model such as SaaS or on-premises.
Setup and configuration
We build the system to match your environment and configure initial settings such as speakers and templates.
Training
We test it under conditions close to your actual meetings and adjust speaker recognition and templates to fit your operations.
Go-live
Operation begins, with settings reviewed and adjusted as needed based on usage.
* Timeframes are estimates and vary depending on requirements.
Use cases
On-site meetings
For on-site meetings where multiple stakeholders gather, decisions and owners are recorded automatically.
Internal team meetings
Automates the creation of minutes for regular and project meetings, shortening the time to sharing.
Client and partner meetings
Accurately captures who said what during negotiations or partnership discussions, ready for later reference.
FAQ
How is this different from YAKUON?
YAKUON's strength is multilingual real-time translation, while Meet AI+'s strength is running meeting minutes in a fully closed environment. You can choose based on your requirements.
What kinds of input are supported?
Audio and video file uploads are supported. Microsoft Teams and calendar integration also let you import meeting recordings directly.
Is conversation data used to train AI models?
Conversation data is not used to train other AI models. The approved LLM runs in a closed environment, handling terms are agreed on a per-project basis, and data is strictly managed as confidential.
Can it handle industry-specific terminology?
It can be tuned to match industry-specific terminology. Terms prone to misrecognition can be registered in advance to improve accuracy.
Can it run in our own environment?
Yes. It can be deployed on-premises or within your own environment, operating to match your information governance requirements. SaaS deployment is also available on request.
Can we try it before deploying?
Yes. A free consultation lets you confirm how it performs under your security requirements on a small scale (an estimate; varies by project).
Operated to match your information governance requirements.
Tell us about your security requirements, and we'll propose an operating model.
* Transcription and summarization accuracy, and the figures shown (creation time, speaker separation accuracy), vary depending on the usage environment (based on our own findings).
* "200+ projects" is the development track record of operator Miichisoft (our own track record).