A clinical decision layer
on top of your HBYS
TIS is Opinion AI's clinical decision platform for the hospital side. The hospital's existing HBYS (hospital information system) stays in place; TIS is added on top as a separate layer and uses the AI agent MINA to interpret HBYS, Medula (the social security pre-authorization and billing system), MKYS (the Ministry of Health's inventory management system) and clinical data flows. Pre-authorization, the digital twin, ICD-10 and SUT (social security reimbursement rules) checks, audit indicators, inventory and medical accounting can all be followed in one place. Data never leaves the hospital, and the decision always rests with people.
The work TIS does
in the hospital
The HBYS keeps the hospital's records; TIS interprets the same data and flags gaps and risks in advance. People always make the decision. AI shows the rationale behind each recommendation and the source it relies on, and every output can be audited later.
Pre-authorization
Patient files are gathered in a single panel, and each application carries a risk score and a status label. The operator moves straight to review without re-entering any data, then approves the application, requests missing documents or rejects it.
Digital twin
With every application, the patient's chronic conditions, allergies, clinical notes, vital signs, lab data and prior visits come together in a single profile. As a result, information from scattered documents can be read in a single clinical context.
Clinical assistant
From a single conversational interface, the AI agent MINA suggests ICD-10 codes, checks SUT compliance and analyzes discharge summaries; it also answers questions about rejection risk, policy coverage and fraud risk. Every answer shows its source and rationale, and when a patient is selected, the output is written to that patient's digital twin record.
Audit cockpit
Before an official audit arrives, the hospital's position is simulated against the BH-13, BH-32, BH-33, BH-34, BH-35 and ADSH-13/14 indicators. Indicators that carry risk are flagged in the management cockpit.
Inventory and MKYS
MKYS data is pulled regularly, and items past 60 and 90 days are listed for the finance director as stock to return or consume. The stock-to-consumption ratio and an action list for critical items are on the same screen.
Medical accounting
SGK (social security) accruals and payments are tracked per application, with collected, pending and rejected amounts visible in real time. Medula billing errors and SUT mismatches are flagged before the invoice is issued.
Data twins for the patient
and the physician
Information about a single patient sits in HBYS and Medula as scattered records in different formats. The digital twin brings these records together on two sides: it builds one live profile for the patient and gives the physician personalized decision support.
Patient twin
HBYS and Medula data merge into one live patient profile. Discharge summaries and reports are read and turned into structured fields.
When physicians open the patient, they find the clinical summary, risk flags and SUT checks in the same view.
Physician twin
The physician twin learns the physician's test-ordering patterns, coding preferences and treatment patterns. This behavioral profile is created with the physician's consent and compared against anonymized peers.
Decision support is personalized on that basis. Recommendations are based on the physician's own practice and are always presented with their rationale.
What changes
in the hospital
The effect of TIS can be seen in the hospital's daily work, particularly in pre-authorization, billing, audit and inventory.
Pre-authorization gets faster
Operators no longer open each file from scratch; they start from the clinical synthesis, risk score and eligibility analysis prepared by MINA.
Fewer rejections
Missing documents and incorrect ICD-10 codes surface before the application reaches the other party. SGK and Medula billing errors are also caught before payment.
Ready for audit
The audit score can be monitored every day, without waiting for the audit date. BH and ADSH indicators are monitored continuously, and those carrying risk are flagged in the cockpit.
Less inventory loss
MKYS data is scanned every day against the 60- and 90-day thresholds. Items at risk of expiring or missing their return window are listed before the loss occurs.
HBYS stays as it is
The hospital's HBYS and workflows remain in place; TIS is added on top of them as a layer that interprets the existing data.
One screen for management
The finance director, chief physician, insurance unit and audit team open the same cockpit instead of pulling separate reports. Because everyone looks at the same data, meetings can focus directly on the actions to take.
Model results on hospital tasks
You can find the task-level evaluation results for MINA v3.7 on the MINA v3.7 page.
Data flow from the HBYS
to the cockpit
Once TIS is connected to the hospital's systems, the steps below run automatically in the background, and the result reaches management as an up-to-date cockpit.
Export from the HBYS
Application, procedure, discharge summary, laboratory and imaging records are received through XML/CSV export or a direct database connection.
Normalization and masking
Data from different sources is merged into one standard schema. Missing fields, format errors and duplicate records are cleaned up, and personal data is masked in line with KVKK (Turkey's personal data protection law).
Writing to the data warehouse
Cleaned records are written to the hospital's own TIS data warehouse, and digital twins and application profiles are updated.
Audit simulation
Active applications are calculated against the BH and ADSH indicators, and an estimated audit score is produced before the official audit.
Gap analysis
Deviations between target and current state are broken down by specialty, case and item, and every indicator that requires action is labeled.
Management cockpit
Critical alerts and recommendations are updated in the cockpit. When the finance director, chief physician and operations team open the same screen, they see the processed output.
Working modules
in the cockpit
TIS starts with a single module, and new modules are added as needed. Alongside pre-authorization, inventory and accounting, day-to-day work runs in the modules below.
Applications
All pre-authorization applications come together in one flow, with live counts of total, under review, pending and approved applications. Each application has its risk score, digital twin link and decision panel ready.
Explore
All applications can be searched by patient name, application number, ICD-10 code or procedure. Results can be filtered by status, risk score and application type.
Statistics
Total applications, approval rate, average risk score and approved amount are visible at a glance. Monthly trends, risk and status distributions and specialty-level analysis sit on the same page.
Documents
Policy, pre-authorization, claim, rejection and appeal, discharge summary, laboratory, MRI, CT and X-ray, pathology, consultation, surgery, prescription, medical board report and ID documents are uploaded by type. PDF, DOC, JPG and PNG files are analyzed automatically by AI, and batch OCR is supported.
ICD-10 coding analysis
The accuracy of each diagnosis code is checked by AI, and the code is classified as "at risk" or "appropriate". An alternative code is also suggested, together with its rationale.
MINA management panel
The cockpit shows the hospital's overall status on a single screen. The audit simulation score, critical indicators, current alerts and a timeline of process status appear together here.
Regulation, data security
and integration
TIS works in line with the SUT and Medula rules of SGK (the social security institution), the Ministry of Health's BH and ADSH indicators and MKYS inventory management. To prepare for audits, it monitors the BH-13, BH-32, BH-33, BH-34, BH-35, ADSH-13 and ADSH-14 indicators. In pre-authorization and billing, it checks SUT compliance, ICD-10 coding accuracy and the SGK and Medula pre-authorization rules, and it tracks MKYS inventory against 60- and 90-day thresholds. Personal data is masked in line with KVKK, and data stays inside the hospital.
Data stays inside the hospital
Data is not transferred outside the hospital. The database is reachable only from the local network, and administrators connect through an SSH tunnel. Connections are encrypted with HTTPS and storage with AES-256. The hospital remains the data controller; Opinion AI does not take institutions' patient data into its own systems.
The existing HBYS is not replaced
Records are transferred automatically by patient number, through XML/CSV export or a direct database connection. The HBYS and its workflows stay in place; TIS is a layer added on top.
On-premise or isolated cloud
TIS runs on an isolated tenant architecture and can be deployed on-premise or in an isolated cloud assigned to the hospital. Modules scale separately: you start with one module and add others as needed.
Related guides and articles
Start with a POC
in your hospital
You can evaluate TIS through a POC that starts with a single module and runs on your hospital's own applications and indicators. Setup and pricing are tailored to your hospital's size and the modules you choose.
POC Request