Want to use AI for meeting recordings, transcripts and meeting minutes, but cannot put recordings on a public cloud? The AIMochi Enterprise On-Premises Edition deploys voice-to-text and meeting notes in your own environment, so recordings and transcripts stay in-house. LargitData installs an offline version on site, on Mac or NVIDIA GPUs, with real-time and batch transcription and an optional in-house LLM.
The on-premises edition suits organizations whose recordings involve confidential or personal information, or are subject to compliance requirements, and that cannot hand their data to an external cloud service. For example:
Whether the edition meets your organization's specific rules is assessed case by case, based on the feature scope, deployment approach and your own policies.
| Cloud version (LINE official account) | Enterprise On-Premises Edition | |
|---|---|---|
| How to start | Add the LINE official account @aimochi and start using it | Contact LargitData for an assessment; we install an offline version on site |
| Where data is stored | The AIMochi cloud service | Your own environment, running offline |
| Hardware | None; LINE on your phone or computer | Mac or NVIDIA GPU environments (including DGX and DGX Spark) |
| Transcription | After you send a voice message or audio file | Real-time and batch transcription, with an optional in-house LLM |
| Cost | Free plan, Basic NT$299/month, Team NT$899/month (Pricing) | Quoted to your needs |
| Best for | Individuals and small teams who want to start right away | Organizations with security, privacy or compliance requirements |
Teams that are still evaluating can first try the cloud version's free points on non-confidential recordings to check how good the transcripts and summaries are, then decide whether to discuss the on-premises edition.
The exact hardware and feature scope are confirmed during the needs assessment, based on the number of users, your recording volume and your use cases.
AIMochi is developed by LargitData. Founded in 2014 and headquartered in Taipei, LargitData has long worked on natural language processing (NLP), large language models (LLM), retrieval-augmented generation (RAG), optical character recognition (OCR) and automatic speech recognition (ASR), serving customers that include government, defense and critical infrastructure organizations. For related technology and products, see LargitData's ASR voice-to-text, AIMochi product introduction and QubicX on-premises AI solution. For more background, see About AIMochi.
1. Contact us: Email [email protected] or call +886-2-2585-5080 and tell us about your organization and needs.
2. Needs assessment: We suggest first gathering the number of expected users, roughly how many hours of recordings you have each month, your main use cases, and the boundaries your data cannot cross along with any security policies you already have.
3. Plan and quote: Based on the assessment, we recommend hardware, define the feature scope and provide a quote, then LargitData installs the system on site.
Yes. AIMochi offers an Enterprise On-Premises Edition that can be deployed in your own environment so recordings and transcripts stay in-house. The feature scope, deployment approach and quote depend on your needs. Contact LargitData by email at [email protected] or by phone at +886-2-2585-5080.
LargitData installs an offline version on site at your organization. It runs on Mac and on NVIDIA GPU environments, including NVIDIA DGX and DGX Spark. Features include real-time and batch speech transcription, and you can add a large language model (LLM) running in-house to turn transcripts into summaries and meeting minutes. The exact hardware and feature scope are assessed based on the number of users and your recording volume.
The cloud version is used through the LINE official account @aimochi: add it as a friend and you can start right away, with the public Free, Basic and Team plans. The on-premises edition is deployed in your own environment so your data stays in-house, and it is discussed and quoted according to your needs.
We suggest gathering the number of expected users, roughly how many hours of recordings you have each month, your main use cases (for example internal meetings, interviews or customer service calls), and the boundaries your data cannot cross along with any security policies you already have, so we can work out a suitable plan.