Chat tools
Where your people ask. Answers come back in the same place.
LINE
Web app
Desktop app
Slack later
Teams later
Email later
Every piece is a plug-in
WenmioPlug in your tools. Get to work.
Wenmio connects the chat app your team already uses, like LINE, to your work systems, files and database, and answers with the AI model you choose. Start with one team and one system, and add more as you go.
- Mac, Windows and Linux
- English and Traditional Chinese
- Runs on your own machines
Orders shipping this week
gpt-oss 20B · on site
Kevin Wu
Which orders ship this week? Flag anything that's late.
The problem
Most AI tools stop at the chat box.
It doesn't know your systems
Staff copy numbers out of the ERP, paste them into a chat window, then copy the answer back.
It lives in yet another app
Your team already talks in LINE. A separate AI app is one more place to remember to check.
The flexible ones need engineers
AI frameworks can do almost anything, once someone builds, connects and runs them for you.
No one can say where the data goes
Questions, files and system data end up on servers nobody at the company chose.
Plug-ins
Every piece plugs in.Start small. Add more as you go.
Wenmio sits in the middle. Your chat app, work systems, data and AI model plug in around it, and each one can change without touching the others.
Work systems
Any system with an MCP server: your ERP, CRM, inventory or an in-house tool. Wenmio looks things up as the person who asked.
Any MCP server
Each person signs in
One company account
Wenmio
Your logo, your people and what each may see, your procedures
Data
Start with the built-in database. When you're ready, point Wenmio at your own database and file storage.
Built-in
Postgres on a NAS
Supabase
S3
NAS shared folder
AI model
Any model with an OpenAI- or Anthropic-style API, hosted or on your own hardware. Switch in Settings at any time.
ClaudeGPTDeepSeekQwengpt-ossLM StudioOllamavLLM
See it at work
Simple for your team.Careful with your data.
Add a system with one form
Anything with an MCP server plugs in: your ERP, CRM, inventory or an in-house tool. Pick how people sign in, and Wenmio finds what it can do.
Ask where your team already talks
Mention Wenmio in a LINE group or message it directly. Anything it wants to post waits for the person who asked.
It reads your files and makes new ones
Upload PDF, Word, Excel and CSV files. Wenmio finds passages by meaning, even when the words differ, and hands back Excel or Word files.
Teach it how your company works
Write a procedure once, in plain words. Wenmio follows it every time the task comes up.
People see only their own data
Wenmio uses each system with the sign-in of the person who asked.
Sensitive systems stay on site
Mark a system on-site only, and Wenmio uses it only with a model on your own hardware.
Your company's name on it
Staff see your name and logo. Wenmio is a small “Powered by Wenmio”.
Two languages, built in
Each person picks English or Traditional Chinese.
How it works
Up and running in three steps
- 01
Install it on your own machine
A PC in your office, a NAS that runs Docker, a Mac or a cloud server. The first visit sets up your company's name, logo and admin.
- 02
Plug in your model, chat app and systems
Each one is a form in Settings or Connectors: a model key or address, your LINE channel keys, a system's address.
- 03
Bring in your team
Each person signs in with their own login and links their own LINE. Then they just ask.
What you'll need
- A computer that stays on: Windows, Mac or Linux
- A NAS, if you want your data kept on it
- For a model on your own hardware, an Nvidia GPU with about 16 GB of memory. Otherwise, a key for a hosted model
- A LINE Official Account, if your team will ask in LINE
Where it runs
You decide where it runs, where your data lives and which model answers.
Pick one of each to see what your setup looks like.
Questions
What companies ask first
Does our data leave the company?
Only if you choose so. With the server, the data and the model on your own machines, nothing leaves your network. With a hosted model, questions and the passages needed to answer them go to that model's provider.
Do we need programmers?
Not to set it up. Chat apps, models and systems that already have an MCP server are added with forms. A system without one needs a small connector built first, and we can help with that.
Which AI model should we pick?
A model on your own GPU keeps everything in house. A hosted model such as Claude needs no new hardware. You can change it in Settings at any time.
Can people see each other's data?
No. Wenmio uses each system as the person who asked, and in a LINE group it never uses anyone's own files.
Do we have to use LINE?
No. Staff can use the web app or the desktop app for Mac and Windows. LINE is one plug-in; Slack, Teams and email are planned.
Which languages does it speak?
The app is in English and Traditional Chinese, and each person picks their own. Wenmio answers in the language it's asked in.
Details
What it runs on and works with
- Languages
- English and Traditional Chinese, chosen by each person
- Chat tools
- LINE; web app; desktop app for Mac and Windows. Slack, Teams and email later
- Work systems
- Any MCP server, with per-person sign-in, personal tokens or one company account
- Server
- Mac, Windows or Linux; a NAS that runs Docker; Nvidia GPU machines; a cloud server
- Database
- Built in, or Postgres with pgvector on your NAS, your server or Supabase
- File storage
- A folder or NAS share, Supabase Storage, or S3
- Models
- Any OpenAI- or Anthropic-style API. Hosted: Claude, GPT, DeepSeek. On your hardware: gpt-oss, Qwen and others through LM Studio, Ollama or vLLM
- Procedures
- Plain-language instructions, one file per task
- Library
- PDF, Word, Excel and CSV, searchable by meaning
- Accounts
- Each company has its own account, name and logo; each member signs in with their own login
- Backups
- Scheduled backups of the database and files
Book a demo
Bring Wenmio to your company
Tell us which chat app and systems your team uses. We'll show you Wenmio working with them, or help you set it up.
hello@wenmio.com