We build the AI transformation of organizations that handle sensitive data such as healthcare and long-term care data.
Where special care-required personal information is handled every day, what comes first is not “using AI” but “a setup that lets you use it safely.” Kurashi Tech supports you end to end, from designing how data is handled through AI adoption to organizational change and staff redeployment, and delivers the business impact of reliably improving profit within the bounds of regulation.
Why Kurashi Tech
① A track record of running operations within the bounds of regulation
For many consulting firms and AI vendors, special care-required personal information in healthcare and long-term care is “a risk to be avoided.” They can draw up the broad plan, but the steps that touch real data are left to the customer or a specialist vendor. Kurashi Tech is the one that has been running those “left-over steps.” We have run the record-keeping and billing work of 200 facilities inside the customer’s own environment, and we hold the returned-claim data. This track record of running operations within regulation is a difference that a plan alone cannot create.
Contractor requirements in Japan’s medical information security guidelines
Special care-required personal information under the APPI
Next-Generation Medical Infrastructure Act
Mandatory cybersecurity measures for medical institutions
Restrictions on taking medical data out of Japan
Typical case
Typical consulting firm
Special care-required personal information tends to become something to avoid
In areas where regulations overlap, they can produce a desk-based plan and an organization chart, but the steps that touch real data are left to the customer or a specialist vendor. The plan gets delivered, yet no one shows up to actually run the records and billing.
Typical case
Typical AI vendor
Installing AI tools tends to become the goal itself
The platform and tools get installed, but meeting contractor requirements, drawing up AI usage rules, and reworking the flow of operations tend to be left to the customer. They often fail to take root on the front line, and it is not unusual for costs to be all that increases.
Kurashi Tech
Running operations within regulation
Real-data experience goes straight into the design
We have run the record-keeping and billing work of 200 facilities inside the customer’s own environment, without the data ever leaving it. We hold returned-claim data and correction histories from Japan’s claims review and payment bodies (the prefectural Federations of National Health Insurance Organizations and the Health Insurance Claims Review & Reimbursement Services), and support you end to end from planning to execution to organizational change and staff redeployment.
200 facilitiesRecord-keeping and billing run inside the customer’s environment
Inside the customer’s environmentOperations run without data leaving the environment
Returned-claim dataWe hold returned claims and correction histories from Japan’s claims review and payment bodies
② Extensive real healthcare data and knowledge
Getting high-quality output from AI requires a large amount of high-quality data. Because Kurashi Tech employs its own domain experts and has itself provided outsourced administrative work for healthcare and long-term care providers, we hold high-quality real data and knowledge in healthcare.
What matters is having high-quality data
Without high-quality data, AI cannot produce high-quality output. Generic data from the internet cannot power AI for real operations. AI becomes usable only when it has real data born on the front line.
Real-world data from our outsourcing work
The record-keeping and billing practice of 200 facilities, the history of returned claims and corrections, and audit-response records. This is data verified on the front line, and it is also put to use in our AI agent series SHISHI and our AI platform CAIVA.
Typical AI projects start by collecting data
Collecting data takes an enormous amount of effort on the client side too, time is consumed sorting out how to handle special care-required personal information, and the project becomes more likely to stall.
Kurashi Tech starts with the data already in place
Because we start with data and operational knowledge already in hand, almost no basic interviews with the client are needed. The scope of data handling can also be defined at the first session, which makes it extremely unlikely that a project will stall.
③ End-to-end service
Unlike consulting services that offer only desk-based discussion, and unlike AI vendors whose only aim is installing tools, a single team supports you through designing how data is handled, the AI platform, reworking operations, rules and audits, and finally organizational change and staff redeployment as the end point.
Data handling design Classification, processing conditions, contractor requirements
Secure AI platform Stored in Japan, not used to train AI models, logs preserved
Reworking operations AI drafts, people decide
Rules and audits AI usage rules and contractor management
Organizational change and redeployment Realizing the business impact
Overview of Kurashi Tech’s AX SolutionsStructure
We target every organization that deals with medical data. In addition to hospital and clinic chains, that means healthcare BPO, healthcare finance, medical trading companies, medical logistics, and pharmaceuticals. In all of them, special care-required personal information flows through operations, and they must operate within the same regulations. A team of PMs and FDEs designs and implements AX applications for each industry, and beneath them sits a common foundation made up of a handling platform for special care-required personal information, an AI agent platform, real healthcare data, rule templates, and a verification platform.
For every organization that deals with medical data
Hospitals
HospitalAX application
Records, billing, bed operations
Clinic chains
Clinic chainAX application
Booking, claims, multi-site operations
BPO
BPOAX application
Claims checks, medical billing services
Finance
FinanceAX application
Receivables review and credit
Trading companies
Trading companyAX application
Ordering of devices and drugs
Logistics
LogisticsAX application
Drug and specimen delivery, temperature control
Pharma
PharmaAX application
Safety data, trials, document control
Project Manager (PM)
Designs strategy, operations and organization
Data inventory
Goal setting
Workflow design
Redeployment design
PM × FDE
Forward Deployed Engineer (FDE)
Builds and works alongside you inside your environment
Requirements
Design and build
Integration
Operational improvement
Special care-required personal information platform
Inside your environment, not used to train AI models, logs preserved
AI agent platform
SHISHI · CAIVA
Real healthcare data
Records, billing and returned claims from 200 facilities
Rules and governance templates
AI usage rules and contractor management
Security and verification platform
Contractor requirements, audit readiness
Kurashi Tech common foundation (the base that runs within the regulations)
How Kurashi Tech thinks about AX SolutionsPhilosophy
There is only one reason a company should adopt AI: “adopting AI raises profit.” Kurashi Tech uses AI as a means and focuses on raising profit. In healthcare and long-term care, however, a single personal-information incident can cost you both profit and trust. That is why Kurashi Tech first builds a setup that can handle special care-required personal information safely, and on top of that supports organizational change and redeployment for the employees whose labor is part of your cost of sales, reliably raising profit.
Principles
Fundamental profit improvement comes from “reducing cost of sales”
Profit includes gross profit and operating profit, but fundamental profit improvement is achieved by reducing cost of sales. AI’s role is to cut the labor costs that make up cost of sales. Our role is to support you through to redeploying and optimally placing the employees replaced by AI.
People should focus on work that raises revenue
People focus on revenue-generating work, starting with building trust. Making that work more efficient is also AI’s role. This work is a strong candidate for where employees replaced by AI can be redeployed.
It is not as simple as “just add AI and costs will fall”
Using AI without a clear plan only raises AI costs and does not improve profit. You need to plan and manage carefully where and how to introduce AI, and how to change your organization so that profit rises.
Keep special care-required personal information in-house and keep operations running
Information about service users and patients is processed inside the customer’s environment, on a platform where the data is not used to train AI models. We define what may be handed over and the processing conditions for each type of data, and keep logs. The precondition is making “protecting” and “using” work together at the same time.
A setup that can run safely Handling design, platform, rules
AI reduces cost of sales Creating surplus staff
Move people to revenue-generating work
Revenue rises
Operating profit improves
Approach
STEP 1: Goal setting and data inventory
We clarify which costs should come down and which revenue should go up in your P&L or management indicators, and break down their causes. At the same time, we take inventory of the data involved, sorting it into special care-required personal information, personal information, anonymously processed information, and operational data, define the target organizations and operations, and set the final goal.
STEP 2: AI vision planning
We make the KGI visible and define the to-be state that achieves the final goal. We set a data handling policy that covers the scope of data that may be given to AI, processing conditions, and retention periods, and clarify the changes required (mainly organizational change and staff redeployment).
STEP 3: Defining measures
We confirm and finalize security requirements such as those in Japan’s government guidelines on medical information security (the “3 Ministries 2 Guidelines”), and select AI that can deliver the to-be operations. We put in place AI usage rules, contractor management, and incident response procedures, and also raise the project team’s AI literacy.
STEP 4: Execution
We develop the AI, deploy it in part of your operations (a small start), and measure its effects and risks. We then draw up a full rollout plan and roll it out across the organization.
STEP 5: Staff redeployment
We change how operations are staffed, redirect the freed-up hours to revenue-generating work, and carry out the placement adjustments. We then track how far the target indicators have been achieved, keep up with changes in law, and run the improvement cycle.
AX starts with goal setting and a data inventory. It continues with AI vision planning, defining measures, and execution, and your management indicators change significantly once you carry it through to staff redeployment at the end.
How projects are staffedProject Team
As a rule, projects are run by a team of just one project manager (PM) and one Forward Deployed Engineer (FDE). Throughout the project, work is organized around a weekly regular meeting, and during the STEP 4 execution (AI development) phase, the FDE can also be stationed on site with you.
Weekly regular meetings throughout the project
STEP 1Goal setting and data inventoryPM
STEP 2AI vision planningPM
STEP 3Defining measuresPMFDE
STEP 4ExecutionPMFDEOn-site / off-site
STEP 5Staff redeploymentPM
Forward Deployed Engineer (FDE): works alongside you inside your own environment and takes charge of everything from identifying issues to implementing the system. The setup lets development and verification proceed without special care-required personal information ever leaving your environment.
Quick Start
So that you can experience firsthand how AI cuts workload, we offer a Quick Start Plan. An FDE visits about one day a week to move AI development forward.
STEP 0 Quick start
STEP 1 Goal setting
STEP 2 AI vision planning
STEP 3 Defining measures
STEP 4 Execution
STEP 5 Staff redeployment
To first see how it works and what real effect it has on the front line, we also offer a plan that carries out part of the AI development and deployment normally done in STEP 4 as STEP 0.
Pick the work you want to try first
Choose the work you want to try first, and we automate it using AI. Even in the Quick Start, we define the scope of data given to AI and the processing conditions at the first session.
An FDE visits about one day a week
With a project manager assigned, an FDE visits about one day a week to move AI development forward inside your own environment.
For IT & DX Departments
Our plan for IT departments and DX departments offers medical institutions credential-stuffing protection, support for adopting general-purpose AI, and vendor selection and vendor management on your behalf.
Free
Free credential-stuffing check
For medical institutions
In response to the notice “Request to Strengthen Security Measures Following Unauthorized Login Incidents in Online Medical Consultation Systems,” issued by the Ministry of Health, Labour and Welfare on October 28, 2026, we offer a free credential-stuffing check. Specifically, we check the effectiveness of multi-factor authentication, password reuse, suspicious logins, and more.
Setting AI usage rules for organizations that handle special care-required personal information
In moving forward with general-purpose AI, we raise the points (issues) that need to be addressed and help resolve them. Specifically, we provide full support for the purpose of adoption, what to have the AI do, the effects you can expect, the general-purpose AI best suited to that work, drawing up usage rules and guidelines, the scope of handling of special care-required personal information, ISMS compliance, and more.
We act on the “buyer’s side” for IT and DX departments
From selecting vendors for AI and system adoption to managing progress after the contract is signed, we act on your behalf from the buyer’s side. We carry out requirements definition and RFP drafting, candidate evaluation based on contractor requirements such as those in Japan’s medical information security guidelines, comparison of quotes and proposals, organizing contract terms, and post-adoption regular meetings, issue management, and acceptance inspection. Drawing on the know-how of running the operations of 200 facilities, we evaluate by whether it works on the front line.