What TOMO does
Let AI run your consultations.
How to take out the garbage, how to find a hospital, how to read a pay slip. Generative AI (large language models) answers workers' questions in 12 languages and follows up until they are resolved.
Answers from an AI that knows each worker
No starting from zero every time.
TOMO answers based on each worker's past consultation history, their host company's information, and the know-how your organization has built up.
Take a request like "Tell me about a hospital nearby." A typical chatbot has to ask for the address every time. TOMO already knows where that worker lives. It also has information on hospitals other workers in your organization have used. So it can give the right guidance from the start.
Behind this knowledge is a three-layer knowledge base.
Base knowledge
Core information shared nationwide: residence status, labor law, social insurance, daily-life manners
Example: "How do I apply for health insurance?"
Area knowledge
City-specific offices, multilingual medical institutions, garbage sorting rules, and more
Example: "How do I take out the garbage in Nagoya?"
Custom knowledge
Know-how unique to your organization, built up automatically from day-to-day support
Example: "For our host companies, we first direct residence certificate requests to the XX office"
For every answer, the AI searches all three layers and combines the information most relevant to that worker. Knowledge unique to your organization takes top priority, so the more you use it, the more answers reflect how your organization does things.
Staff step in when judgment is needed
Not everything is left to AI.
For informational questions such as how to take out the garbage, the AI answers automatically and that's the end of it. But not every consultation is like that.
When an outside agency must be contacted, when paperwork is involved, or when a worker asks to speak with staff, a person needs to decide. In these cases, the AI drafts a response and shows it to staff on the dashboard.
Staff review the content and simply press "Approve and send." If changes are needed, they can edit it on the spot before sending. The AI never contacts outside parties on its own.
To handle any kind of consultation consistently, the AI has 13 built-in response templates.
As soon as a worker's message arrives, the AI automatically picks the best template. Every template follows the same four steps: listen, provide information, plan outside action if needed, and follow up.
Follow-up until it is resolved
No consultation is left hanging.
Answering a question is not the goal. TOMO's consultations see it through to what happened afterward.
After a set time, the AI checks in with the worker: "How is it going?" If it's resolved, the worker replies "Resolved" and that's the end. If not, the AI automatically reworks the response plan and suggests the next step.
When a consultation is complete, what happened and how it was resolved are recorded automatically as a case summary. This record is added to your organization's knowledge, so the AI handles similar consultations even better next time.
TOMO covers the main languages of workers in Japan. Text where accuracy matters most, such as legal consent documents and emergency contacts, is translated in advance. Everyday consultations are translated by the AI in real time, so conversations feel natural.