We define your institution's problem with you,and build all the way through.
Adoption and collaboration start with looking at your field together. From discovery to operation, we grow it with you, watching how it is actually used.
Docspert — answers based on your regulations.
An evidence-based AI for regulations and documents. It finds the basis in your regulations, guidelines and bylaws to answer, and attaches a source down to the article and clause to every answer.
Example question: By when do I need to settle travel expenses after a business trip? Example answer: The expense report is due within 14 days of the day the trip ends. Receipts are attached to the report, and amounts paid by corporate card can be replaced by the payment record. If the deadline is missed, a confirmation from the department head is required. Sources: Travel Expense Regulation Art. 12(1), Accounting Guidelines 3.2, Travel Expense Regulation Art. 12(3).
Ask, and it answers with its basis
Ask a question and it finds the relevant provisions to answer, marking each sentence with the regulation, article and clause it comes from.
It flags the provisions to revise first
The regulation check-up finds outdated provisions and provisions that conflict with each other, and lists them for you.
It writes drafts with their basis attached
It drafts notices and official letters in line with your regulations, and records what each part is based on.
UniQ — many models, one account.
A generative AI portal for organizations. Choose the model that fits the job, pay only for what you use, and keep a record of which company each member's words are sent to.
Example: when the Planning team sends "Summarize the meeting minutes in three lines", UniQ picks the fast domestic model that suits a short summary, replies with a three-line summary, then adds "Planning team · Meeting minutes summary · Domestic" to the destination log and raises the Planning team's usage.
It picks the model that fits the job
Short summaries go to a fast model, long documents to a model suited to them. You keep one account.
You pay only for what you use
See at a glance how much each team has used, and pay only for that.
It records where each sentence was sent
It records which team's words went to which company's model.
From discovery to operation, all the way through.
- 01
Discover
We observe the field, the data and the users closely to define the real problem. We start from the problem, not the technology.
- 02
Design
From a small first version to a structure that can grow. We design the AI and data pipelines together with the user experience.
- 03
Build
We build fast and test where it is actually used. A working product proves the hypothesis.
- 04
Operate
Deployment, monitoring and improvement. We stay responsible until the value reaches users.
When you first contact us
It is fine if you are not fully prepared. We will sort it out together in the first conversation.
- The problem you face and its situation
- Examples of related documents or data (as far as you can share)
- A contact person to check things with
Frequently asked questions
Can a single department or a small organization work with you?
Yes. We recommend starting with one problem and one department. We check small, then widen.
It is hard for us to send data outside.
We first listen to your security requirements, then decide together on a setup that works within them.
How are the timeline and cost decided?
It depends on the scope of the problem. We agree on the scope in the first conversation, then give you an estimate.
What happens after it is built?
Deployment is not the end. We fix it as we watch how it is used, and help run it as long as needed.