RPA implementation and support

RPA

RPA implementation and support

MedSync staff working together around a computer

OVERVIEW

Let robots handle repetitive data entry, transfer, and tallying overnight

RPA (robotic process automation) uses software robots to carry out the routine operations people perform on a computer. The robots work in your electronic health record (EHR) and Excel, following steps set in advance, and process the work in batches at the times you choose.

MedSync supports medical institutions from workflow analysis through robot development, deployment, operation, and improvement.

When on, the diagrams no longer advance on their own or animate. Choosing an item still switches the view.

SampleWhich tasks, in which departments and facilities, can robots take on?
Choose a department or facility
  • Every day18:30Sending finalized documents by fax or through the regional care-coordination platformBy: RPAExampleCommunity liaison officeWhat robots and AI doSend finalized documents to each recipient.What you checkStaff review a list of anything not sent and follow up.Already deployed for the same operation at a clinic that provides home medical care (faxing home care guidance reports to care managers)See it in action (sending finalized documents by fax or through the care platform)
  • Every day19:00Drafting discharge summariesBy: RPAAIExampleWardsWhat robots and AI doGather the records of patients due for discharge and draft the summaries.What you checkThe physician reviews, edits, and finalizes each summary.See the M-Automate page (drafting discharge summaries)
  • Every day19:30Sending home care guidance reportsBy: RPADeployedClinicWhat robots and AI doSend the day’s guidance reports to each recipient.What you checkStaff review a list of anything not sent and follow up.To have the reports drafted as well, we add AI.See it in action (sending home care guidance reports)
  • Every day20:00Drafting discharge support plansBy: RPAAIExampleCommunity liaison officeWhat robots and AI doGather the records and draft the plan.What you checkThe staff member in charge reviews, edits, and finalizes the plan.See the M-Automate page (drafting discharge support plans)
  • Monthly21:00Preparing home-visit nursing ordersBy: RPADeployedClinicWhat robots and AI doUse the previous month’s orders to prepare and save the next month’s.What you checkThe physician reviews the orders.For now, the robot prepares these orders (the physician’s written authorization for home-visit nursing) on its own. With M-Automate, we’re working on having AI draft the condition field.
  • Month start21:00Downloading and printing invoicesBy: RPAExamplePlanning and adminWhat robots and AI doDownload the invoices, then save and print them.What you checkThe staff member in charge reviews them.
  • Every day22:00Formatting nursing recordsBy: RPAAIDeployedWards, home-visit nursing stationsWhat robots and AI doTidy the records into readable entries and update the EHR.What you checkThe nurse reviews and finalizes them.Deployed at a home-visit nursing station (formatting home-visit nursing records)See the M-Automate page (formatting nursing records)
  • Month-end22:00Extracting and compiling outpatient data submission add-on itemsBy: RPAAIDeployedBilling department, clinicsWhat robots and AI doPick out eligible items from the records and add them to the list.What you checkThe staff member in charge checks the list before submitting it.Deployed at clinics and hospitals that claim this add-on (a fee-schedule item for submitting outpatient care data to Japan’s Ministry of Health, Labour and Welfare)See it in action (extracting and compiling outpatient data submission add-on items)
  • Monthly22:30Payment reconciliationBy: RPAExampleBilling department, planning and adminWhat robots and AI doOrganize the payment data into a list for review.What you checkThe staff member in charge checks the list.
  • Monthly23:00Calculating length of stay by bed typeBy: RPAExampleBilling departmentWhat robots and AI doCalculate the length of stay for each bed type and compile the results into a list.What you checkThe staff member in charge checks the list.
  • Mornings6:00Checking inpatient prescriptions about to run outBy: RPAExamplePharmacy departmentWhat robots and AI doList prescriptions nearing their end date, ward by ward.What you checkPhysicians, pharmacists, and nurses check the list, and the physician makes the prescribing decision.
  • Mornings6:00Alerts about routine test ordersBy: RPAExampleWardsWhat robots and AI doOrganize the information that needs checking and send an alert.What you checkPhysicians and nurses check it and decide what to do.
  • Mornings6:30Daily tally of patient visits and fee-schedule pointsBy: RPADeployedClinicWhat robots and AI doTally the previous day’s patient visits and points, and share the results.What you checkThe clinic director or staff member in charge checks the numbers.See it in action (daily tally of patient visits and fee-schedule points)
  • Mornings6:30Daily tally of patient visits and fee-schedule pointsBy: RPAExamplePlanning and adminWhat robots and AI doTally the previous day’s patient visits and points, and share the results.What you checkThe staff member in charge checks the numbers.Already deployed for the same operation at a clinicSee it in action (daily tally of patient visits and fee-schedule points)
  • Month-end10:00Preparing home-visit nursing reportsBy: RPAAIDeployedHome-visit nursing stationWhat robots and AI doDraft the clinical course from the records and prepare the PDF.What you checkThe nurse reviews and edits the report, then sends it to the attending physician.An example of running the robot during the day.See it in action (preparing home-visit nursing reports)

RPARobot operationsAIReading records and drafting (M-Automate)PersonReview and editing

Deployed: Tasks we have deployed that are now in operationIn progress: Tasks in development or in pilot operationExample: Examples of tasks we can handle (steps are designed for each facility)

If a task follows set steps, we’re happy to talk about it, even if it isn’t listed here.

Billing department: In the billing department (which handles reception, billing, and insurance claims), robots can tally add-on data and prepare review lists overnight.

Community liaison office: In the community liaison office (which coordinates referrals, admissions, and discharges with outside providers), robots and AI can send finalized documents and draft discharge support plans.

Wards: On the wards, robots and AI can draft discharge summaries and format records overnight, so the morning starts with review.

Pharmacy department: The pharmacy department can check prescriptions nearing their end date on a list every morning. Prescribing decisions stay with the physician.

Planning and admin: Every morning at a set time, planning and admin staff get the previous day’s patient visits and fee-schedule points (the unit used to bill medical services in Japan). Robots can also download invoices and reconcile payments.

Clinic: At clinics, robots can send guidance reports and prepare home-visit nursing orders at night, outside consultation hours. All that’s left for the morning is reviewing the results and following up on anything not sent.

Home-visit nursing station: At home-visit nursing stations (agencies whose nurses care for patients at home), robots and AI draft reports and format records. All that’s left for nurses is reviewing and editing the content.

How this diagram works

Choose a department or facility to show only the rows for its tasks. All the row text is written into the page, so the same content appears when you make a choice, even where scripts don’t run.

radio + :has() (filtering without scripts) / View Transitions (reordering rows) / CSS animation (time tags and done markers) / @container (switching row layouts)

In older browsers, all rows are shown, each labeled with its department or facility.

*Times, frequencies, and task order are examples. Actual run times are set to fit each task and workflow.

Supported tasks by type 18 tasks

These are tasks we have already automated, plus examples of tasks we can handle. We also help lighten the admin work required to claim add-ons (extra fee items in Japan’s medical fee schedule), such as preparing items for the data submission add-ons.

Preparing documents7 tasks

  • Home care guidance reportDeployed
  • Home-visit nursing reports and ordersDeployed
  • Home care planIn progress
  • Discharge support plansExample
  • Discharge summariesExample
  • Incident reportsExample
  • Chart entry formattingDeployed

Tallying and submission4 tasks

  • Preparing items for the outpatient and home care data submission add-onsDeployed
  • Length of stay by bed typeExample
  • Daily tally of patient visits and fee-schedule pointsDeployed
  • Analyzing and sharing management KPIsDeployed

Checks and admin work7 tasks

  • Claims checksExample
  • Alerts about routine test ordersExample
  • Checking inpatient prescriptions about to run outExample
  • Unreviewed findings listExample
  • Identifying at-risk patientsExample
  • Downloading and printing invoicesExample
  • Payment reconciliationExample

Deployed: Tasks we have deployed that are now in operationIn progress: Tasks in development or in pilot operationExample: Examples of tasks we can handle (steps are designed for each facility)

*Steps and systems differ from facility to facility, so we build each robot to fit your facility and task.

IN ACTION

Watch a robot work the screens, one step at a time

Choose a task, and the robot opens the EHR and other screens and works through the steps in order. Steps where a person checks the work have a yellow tag. You can also stop autoplay and go through it one step at a time.

SampleFor the task you choose, which screens does the robot work in, and in what order?
1Choose a task

1 Sign in to the EHR

Sakura Home Care Clinic (fictional) · Admin staffTue, Sep 8, 19:30Robot running

EHR (sample)Sign in

User IDrobot-01Robot accountPassword●●●●●●●●Sign in

2 Select what to send from the day’s records

Sakura Home Care Clinic (fictional) · Admin staffTue, Sep 8, 19:31Robot running

EHR (sample)

PatientsClinical documentsAdmin documents
Date range2026/09/08 – 2026/09/08Search
PatientVisit typeSend
Patient A (fictional)Scheduled visitYes
Patient B (fictional)Scheduled visitYes
Patient C (fictional)Unscheduled house callNo
Patient D (fictional)Scheduled visitYes
Patient E (fictional)Scheduled visitYes

Example rule: send reports for patients on scheduled visits

3 Fax to the registered recipients

Sakura Home Care Clinic (fictional) · Admin staffTue, Sep 8, 19:33Robot running

Fax (sample)1 of 4

DocumentHome care guidance report (Patient A, Sep 8)ToKamome Care Management Office (fictional)RegisteredFax numberRegistered number (hidden)SendSent

4 Log unsent reports on a list

Sakura Home Care Clinic (fictional) · Admin staffTue, Sep 8, 19:41Robot running

Spreadsheet (sample)Unsent reports (Sep 8)3 of 4

PatientReasonStatus
Patient D (fictional)No recipient office registeredPending

5 Log the run and report completion

Sakura Home Care Clinic (fictional) · Admin staffTue, Sep 8, 19:53Robot running

Run log (sample)

19:30Start

19:31To send: 4

19:41Patient D: no recipient registered; added to list

19:53End — Sent: 3 · Not sent: 1 · Errors: 0

In-house chat (sample)

Robot (RPA)19:53

Sending is complete. One patient needs your attention. The list has been saved.

Send

6 Next morning, check the list and follow up

Sakura Home Care Clinic (fictional) · Admin staffWed, Sep 9, 8:40Human review

Morning check (admin staff)By: Admin staff

  • ☑ Check the run log
  • ☐ Confirm the recipient for Patient D (fictional) and send manually

Step 1: The robot signs in to the EHR with its dedicated robot account.

Step 2: The robot sets the date range to today and, following your facility’s rules, selects the patients on scheduled visits whose reports will be sent.

Step 3: The robot faxes each report using the registered recipient and number, repeating for every patient in scope. It can also upload the reports to the care platform.

Step 4: If a patient has no registered recipient, the robot logs the patient on a list and moves on to the next one without stopping.

Step 5: The robot logs how many reports were sent and how many weren’t, then lets your staff member in charge know it has finished.

Step 6: The next morning, your staff member in charge follows up on the one patient left on the list. The other three reports went out overnight.

2Choose a stepClick a step to switch to its screen.

To watch again, turn on Autoplay.

Compare this task before and after

All steps (6)
  1. By: RPASign in to the EHR
  2. By: RPASelect what to send from the day’s records
  3. By: RPAFax to the registered recipients
  4. By: RPALog unsent reports on a list
  5. By: RPALog the run and report completion
  6. By: Admin staffNext morning, check the list and follow up

1 Find this month’s patients due for reports

Minato Home-Visit Nursing Station (fictional)Wed, Sep 30, 10:00Robot running

Spreadsheet (sample)Patient list: September

PatientTypeThis month’s report
Patient B (fictional)ContinuingIn scope
Patient F (fictional)NewNurse writes
Patient H (fictional)Not assignedNot needed
Patient J (fictional)ContinuingIn scope

2 Retrieve this month’s nursing records

Minato Home-Visit Nursing Station (fictional)Wed, Sep 30, 10:02Robot running

EHR (sample)

Chart no.000482Search

Patient B (fictional) · Age 81 · Male · Chronic heart failure, hypertension

Nursing records

Tue, Sep 1: BP 128/74, P 76, SpO2 96%, weight 48.2 kg. Lower-leg edema (±). No shortness of breath. Medications checked, no missed doses

Tue, Sep 8: BP 138/80, P 84, SpO2 95%, weight 49.0 kg (+0.8 kg from last week). Lower-leg edema (+). Short of breath on stairs. Daughter: “He’s been eating more store-bought prepared food.” Reported to attending physician by phone; temporary diuretic increase ordered

Tue, Sep 15: BP 126/72, P 78, SpO2 96%, weight 48.4 kg. Lower-leg edema (±). Discussed salt restriction with patient and daughter

Tue, Sep 29: BP 124/70, P 74, SpO2 97%, weight 48.1 kg. No edema. No shortness of breath. Walks steadily indoors using handrails

3 Draft the “Clinical course” section

Minato Home-Visit Nursing Station (fictional)Wed, Sep 30, 10:03AI drafting

M-Automate (AI)

Nursing records (September)

Tue, Sep 1: BP 128/74, P 76, SpO2 96%, weight 48.2 kg. Lower-leg edema (±). No shortness of breath. Medications checked, no missed doses

Tue, Sep 8:BP 138/80, P 84, SpO2 95%, weight 49.0 kg (+0.8 kg from last week)1. Lower-leg edema (+). Short of breath on stairs2. Daughter: “He’s been eating more store-bought prepared food.” Reported to attending physician by phone; temporary diuretic increase ordered3

Tue, Sep 15: BP 126/72, P 78, SpO2 96%, weight 48.4 kg. Lower-leg edema (±). Discussed salt restriction with patient and daughter

Tue, Sep 29:BP 124/70, P 74, SpO2 97%, weight 48.1 kg. No edema. No shortness of breath4. Walks steadily indoors using handrails

Clinical course (draft)Draft

On September 8, weight was up 0.8 kg from the previous week, with lower-leg edema and shortness of breath on stairs. Reported to the attending physician and received an order to temporarily increase the diuretic. The family said the patient had been eating more store-bought prepared food, so salt restriction was discussed with the patient and family. Since then, weight has stayed at 48.1–48.4 kg, and lower-leg edema and shortness of breath have resolved. Blood pressure has been 124–138/70–80 mmHg, and SpO2 95–97%. At the September 1 medication check, there were no missed doses. …

4 Create the report and export a PDF

Minato Home-Visit Nursing Station (fictional)Wed, Sep 30, 10:04Robot running

EHR (sample)Home-visit nursing report

Report dateSep 30, 2026Visit monthSeptemberAge81Clinical courseOn September 8, weight was up 0.8 kg from the previous week, with lower-leg edema and shortness of breath on stairs. Reported to the attending physician and received an order to temporarily increase the diuretic. The family said the patient had been eating more store-bought prepared food, so salt restriction was discussed with the patient and family. …Export PDFSaved VisitNursingReport_PatientB_2026-09.pdf

5 Mark the list and log the results

Minato Home-Visit Nursing Station (fictional)Wed, Sep 30, 10:05Robot running

Spreadsheet (sample)Patient list: September

PatientTypeThis month’s report
Patient B (fictional)Continuing✓ Created
Patient F (fictional)NewNurse writes
Patient J (fictional)Continuing✓ Created

Run results (sample)

Patient B (fictional)Success

Patient J (fictional)Success

Patient F (fictional)Not in scope (new patient)

6 Nurse reviews, edits, and sends the report

Minato Home-Visit Nursing Station (fictional)Wed, Sep 30, 11:30Human review

Report review (nurse)By: Nurse

Clinical course (from line 3)

Since then, weight has stayed at 48.1–48.4 kg, and lower-leg edema and shortness of breath have resolvedno lower-leg edema or shortness of breath was observed at the September 29 visit.

Blood pressure has been 124–138/70–80 mmHg, and SpO2 95–97%.

At the September 1 medication check, there were no missed doses. Next month, we will continue to monitor mainly for changes in weight and lower-leg edema.

  • ☑ Review and edit the content
  • ☐ Send to the attending physician

Step 1: The robot picks out this month’s patients due for reports from the patient list. Reports for new patients are written by a nurse.

Step 2: The robot searches for each patient by medical record number and retrieves this month’s nursing records.

Step 3: AI (M-Automate) handles this step. It reads the records and writes a draft in the station’s own documentation style.

Step 4: The robot builds this month’s report from the previous one. It updates the report date, age, and other fields, enters the draft, then exports a PDF.

Step 5: The robot marks each finished patient on the list and logs which reports were created and why any patient was excluded.

Step 6: A nurse reviews the report, edits it as needed, and sends it. The strikethrough and yellow highlights are the nurse’s edits.

2Choose a stepClick a step to switch to its screen.

To watch again, turn on Autoplay.

See how the draft is written on the M-Automate page

All steps (6)
  1. By: RPAFind this month’s patients due for reports
  2. By: RPARetrieve this month’s nursing records
  3. By: AIDraft the “Clinical course” section
  4. By: RPACreate the report and export a PDF
  5. By: RPAMark the list and log the results
  6. By: NurseNurse reviews, edits, and sends the report

1 Retrieve the month’s outpatient records

Billing department (sample)Wed, Sep 30, 22:00Robot running

EHR (sample)Outpatient visit records

MonthSeptember 2026Retrieve records
PatientSpecialtyRecords
Patient K (fictional)Internal medicineRetrieved
Patient L (fictional)Internal medicineRetrieved
Patient M (fictional)OrthopedicsRetrieved
Patient N (fictional)Internal medicineRetrieving

+38 more

2 Pick out add-on-eligible items

Billing department (sample)Wed, Sep 30, 22:06AI drafting

M-Automate (AI)

Medical record (Patient K, text omitted)

Visit date: Sep 29, 2026 · Internal medicine

Chief complaint and history (text omitted)

Physical exam1 (text omitted)

Test results2 (text omitted)

Prescriptions (text omitted)

Treatment plan3 (text omitted)

Extracted items (Patient K)

1Eligible item · Source: physical exam2Eligible item · Source: test results3Eligible item · Source: treatment plan

Patient K (fictional): 3 items

3 Add the items to the submission list

Billing department (sample)Wed, Sep 30, 22:07Robot running

Spreadsheet (sample)Submission list: September3 of 42

Visit datePatientSpecialtyItemsReview
9/2Patient L (fictional)Internal medicine2Unreviewed
9/15Patient M (fictional)Orthopedics1Unreviewed
9/29Patient K (fictional)Internal medicine3Unreviewed

4 Log the run and report completion

Billing department (sample)Wed, Sep 30, 23:10Robot running

Run log (sample)

22:00Start

22:06Patient K: items extracted

23:10End — In scope: 42 · Added to list: 42 · Errors: 0

In-house chat (sample)

Robot (RPA)23:10

The September submission list has been saved. Please review it before submitting.

Send

5 Staff review and submit the list

Billing department (sample)Before submissionHuman review

Pre-submission review (billing)By: Admin staff

  • ☐ Review the list
  • ☐ Submit

The list is a draft for staff to review.

Step 1: On the last night of the month, the robot retrieves that month’s outpatient records for each eligible patient.

Step 2: AI (M-Automate) handles this step. It reads the text of the records and picks out the items eligible for the add-on.

Step 3: The robot adds the extracted items to the submission list, patient by patient.

Step 4: The robot logs how many records it processed and lets your staff member in charge know the list has been saved.

Step 5: The list is a draft. The staff member in charge reviews it before submitting. Daily manual entry and the month-end sweep give way to a review before submission.

2Choose a stepClick a step to switch to its screen.

To watch again, turn on Autoplay.

Compare this task before and after

All steps (5)
  1. By: RPARetrieve the month’s outpatient records
  2. By: AIPick out add-on-eligible items
  3. By: RPAAdd the items to the submission list
  4. By: RPALog the run and report completion
  5. By: Admin staffStaff review and submit the list

1 Export the previous day’s billing data

Clinic · Director (sample)Fri, Sep 25, 6:30Robot running

Billing system (sample)Data export

Period2026/09/24FormatCSVExport

2 Tally patients and points by specialty

Clinic · Director (sample)Fri, Sep 25, 6:31Robot running

Spreadsheet (sample)Daily tally: Sep 24

SpecialtyPatientsPoints
Internal medicine8641,230
Orthopedics4218,960
Total12860,190

3 Send by email or in-house chat

Clinic · Director (sample)Fri, Sep 25, 6:32Robot running

In-house chat (sample)

Robot (RPA)6:32

Tally for Thu, Sep 24: 128 patient visits, 60,190 fee-schedule points. The breakdown by specialty is attached.

Send

4 Clinic director and staff check the numbers

Clinic · Director (sample)Fri, Sep 25, 8:30Human review

Morning checkBy: Clinic director

  • ☐ Check the previous day’s patient count and points

Step 1: The robot exports the previous day’s billing data from the billing system.

Step 2: The robot tallies patients and points by specialty.

Step 3: At a set time, the robot sends the results by email or in-house chat.

Step 4: The clinic director and your staff member in charge can check the previous day’s numbers first thing in the morning.

2Choose a stepClick a step to switch to its screen.

To watch again, turn on Autoplay.

All steps (4)
  1. By: RPAExport the previous day’s billing data
  2. By: RPATally patients and points by specialty
  3. By: RPASend by email or in-house chat
  4. By: Clinic directorClinic director and staff check the numbers
How this diagram works

The screen mockups are built from text and shapes, not video or screenshots. The text stays sharp, and you can pause on each step to read it.

radio + :has() (switches steps without scripts) / CSS animation (cursor, typing, and check marks) / View Transitions (switching tasks) / IntersectionObserver (autoplays only while in view)

On devices set to reduce motion, the diagram does not autoplay. Click a step to switch to its screen. In older browsers, all screens are shown in step order.

*These samples are based on robots in operation and the steps in their specifications, with the screens and names replaced by fictional ones. Actual steps and screens vary by facility and system. In the home care guidance report example, the robot sends guidance reports that have already been written. To have the reports drafted as well, we add AI (M-Automate). In the outpatient data submission add-on example, the item names are hidden.

*The daily tally of fee-schedule points is a sample based on a robot in operation at a clinic. For hospitals, we design how the tally is broken down (by department, by specialty, and so on) to fit your facility’s workflow.

See how AI reads records and writes drafts on the M-Automate page

After adoption, people review and follow up

Toggle a task between before and after to see which steps move to robots and AI, and what’s left for people.

SampleHow does the number of steps done by people change after adoption?
1Choose a task
2

Steps done by people

Before72After11

Sample based on the manual process at a clinic that provides home medical care and the steps in the robot’s specifications.

Clinics and hospitals that claim the outpatient data submission add-on | Admin staff | Monthly

Time spent tallying and preparing submission data (per month)

BeforeAbout 60 hours

AfterAbout 1 hour

*Results at a client site (using RPA and AI together). Actual results vary by facility, task, and workflow.

Staff now mainly check the submission data instead of tallying it by hand.

Faxing home care guidance reports: before

By: Admin staffSign in to the EHROpen the EHR and sign in.
By: Admin staffSearch today’s recordsSet the document search period to today.
By: Admin staffCheck who to send toCheck whether each patient is on scheduled visits.
By: Admin staffOpen the guidance reportRepeatOpen today’s guidance report for each patient.
By: Admin staffCheck the recipientRepeatCheck that the patient’s care management office is registered as a recipient.
By: Admin staffSend the faxRepeatConfirm the recipient and number, then send.
By: Admin staffCheck for unsent reportsAfter sending, check whether any reports could not be sent.

Faxing home care guidance reports: after

By: RPASign inSigns in with a dedicated robot account.
By: RPASearch today’s recordsSets the search period to today.
By: RPASelect who to send toSelects patients using criteria your facility sets.
By: RPAOpen the guidance reportRepeatOpens the report for each patient.
By: RPACheck the recipientRepeatChecks that the recipient is registered.
By: RPASend the faxRepeatSends to the registered number.
By: RPALog and notifyLists patients whose reports could not be sent, then sends a completion notice.
By: Admin staffReview the list and follow upNewThe next morning, follow up on the patients left on the list.

Extracting and compiling outpatient data submission add-on items: before

By: Admin staffEnter eligible itemsDailyManually enter the required items into a dedicated tab or tool.
By: Admin staffOrganize the data at month-endCheck eligible patients and prepare the data for submission.

Extracting and compiling outpatient data submission add-on items: after

By: RPARetrieve medical recordsPulls the medical records used for the tally.
By: AIExtract eligible itemsNewPicks out add-on-eligible items from the records.
By: RPABuild the submission listCompiles the extracted information into a list for submission.
By: Admin staffReview and submitNewReview the list, then submit it.

Faxing home care guidance reports, before: Before adoption, a medical administrative staff member does all seven steps and repeats three of them for every patient.

Faxing home care guidance reports, after: After adoption, the robot carries out the seven steps. What is left for people is reviewing the list and following up on patients whose reports could not be sent.

Extracting and compiling outpatient data submission add-on items, before: Before adoption, medical administrative staff enter the eligible items by hand every day, then check eligible patients at month-end and prepare the data for submission.

Extracting and compiling outpatient data submission add-on items, after: The robot retrieves the records and builds the submission list, and AI extracts the eligible items. The staff member in charge reviews the list and submits it.

How this diagram works

Cards for the same step share the same name, so when you switch views, the browser automatically animates each card to its new position.

View Transitions (view-transition-name) / radio and checkbox + :has() / CSS Grid

In browsers that don’t support this, the switch isn’t animated, but the content is the same. In older browsers, the before and after diagrams are stacked vertically.

*Samples based on case studies and the steps of robots in operation. The way steps are counted is for illustration only and varies by facility.

See before and after for guidance and nursing reports on the M-Automate pageTalk to us about automating this task

HOW IT WORKS

Works across systems, following your steps exactly

The robot works across your EHR, regional care-coordination platforms, and in-house Excel and Word files. It takes care of everything from retrieving information to entering and saving it. Because it operates screens just as a person does, it can sometimes automate work even in systems that have no integration features.

RPAExamples of screen operations the robot takes over

  1. 1Open the EHR
  2. 2Search for eligible patients
  3. 3Retrieve the records needed
  4. 4Enter data in Excel
  5. 5Save a PDF
  6. 6Enter data into another system
  7. 7Log the results

We design for exceptions and errors, too

Even for the same document, the patients in scope and the processing conditions differ from facility to facility. Before we design a robot, we ask you about the four points below, including what it should do when an exception or error occurs.

SampleWhat does the robot do when something unusual happens?

What we check before designing

  1. Who is in scope?Applies here
  2. What are the processing conditions?Applies here
  3. Where does a person check the work?Applies here
  4. What does it do if an error occurs?Applies here
Choose a scenario

On a normal night

Sakura Home Care Clinic (fictional) · Admin staffTue, Sep 8

Tonight’s runPatients A–E (fictional)

  1. Sign-in: successful
  2. Patient A: sent
  3. Patient B: sent
  4. Patient C: not in scope
  5. Patient D: sent
  6. Patient E: sent

Run log (sample)

19:30Start

19:31To send: 4 (not in scope: 1)

19:33–19:44Sent reports for 4 patients

19:45End — Sent: 4 · Not sent: 0 · Errors: 0

Notification to staff in charge (sample)

Robot (RPA)19:45

Sending is complete. No reports failed to send.

Who does what

  • Robot (RPA)Sends reports based on the patients and conditions you set, then logs the results
  • Staff in chargeChecks the notification the next morning
  • MedSyncNo action needed in this case

If a patient’s report can’t be sent

Sakura Home Care Clinic (fictional) · Admin staffTue, Sep 8

Tonight’s runPatients A–E (fictional)

  1. Sign-in: successful
  2. Patient A: sent
  3. Patient B: sent
  4. Patient C: not in scope
  5. Patient D: added to list
  6. Patient E: sent

Run log (sample)

19:30Start

19:31To send: 4 (not in scope: 1)

19:41Patient D: no recipient registered; added to list

19:53End — Sent: 3 · Not sent: 1 · Errors: 0

Notification to staff in charge (sample)

Robot (RPA)19:53

Sending is complete. One patient needs your attention. The list has been saved.

Who does what

  • Robot (RPA)Adds the patient to the list and moves on to the next one
  • Staff in chargeChecks the list the next morning and follows up
  • MedSyncNo action needed in this case

If the run stops partway

Sakura Home Care Clinic (fictional) · Admin staffTue, Sep 8

Tonight’s runPatients A–E (fictional)

  1. Sign-in: successful
  2. Patient A: sent
  3. Patient B: sent
  4. Patient C: not in scope
  5. Patient D: stopped, then rerun and sent
  6. Patient E: sent

Run log (sample)

19:30Start

19:37Stopped while processing Patient D

19:39Stop detected; rerunning (skipping the 2 already sent)

19:47End — Sent: 4 · Not sent: 0 · Errors: 1 (completed on rerun)

Notification to staff in charge (sample)

Robot (RPA)19:47

The run stopped once but was rerun, and sending is complete.

Who does what

  • Robot (RPA)Detects the stop and reruns, skipping reports already sent
  • Staff in chargeChecks the notification
  • MedSyncNo action needed in this case

If the screen changes

Sakura Home Care Clinic (fictional) · Admin staffTue, Sep 8

Tonight’s runPatients A–E (fictional)

  1. Sign-in: successful
  2. Patient A: run stopped
  3. Patient B: not processed
  4. Patient C: not in scope
  5. Patient D: not processed
  6. Patient E: not processed

Run log (sample)

19:30Start

19:32Send button not found; run stopped

19:32Staff notified; sending not completed

Notification to staff in charge (sample)

Robot (RPA)19:32

The screen has changed, so the run was stopped. Tonight’s reports have not been sent.

Who does what

  • Robot (RPA)Stops the run and sends a notification
  • Staff in chargeHandles the unsent reports and contacts MedSync
  • MedSyncModifies the robot to match the new screen

If the password has expired

Sakura Home Care Clinic (fictional) · Admin staffTue, Sep 8

Tonight’s runPatients A–E (fictional)

  1. Sign-in: failed (run not started)
  2. Patient A: not processed
  3. Patient B: not processed
  4. Patient C: not in scope
  5. Patient D: not processed
  6. Patient E: not processed

Run log (sample)

19:30Start

19:30Sign-in failed (password expired)

19:30Staff notified; run not started

Notification to staff in charge (sample)

Robot (RPA)19:30

Sign-in to the EHR failed, so the run has not started. Please update the password for the dedicated robot account.

Who does what

  • Robot (RPA)Sends a notification without starting the run
  • Staff in chargeUpdates the dedicated robot account’s password and handles the unsent reports
  • MedSyncNo action needed in this case

On a normal night: The robot finishes sending based on the patients and conditions your facility has set, then logs the results and sends a notification.

If a patient’s report can’t be sent: The robot adds that patient to a list and moves on to the next patient without stopping. Your staff member in charge follows up the next morning.

If the run stops partway: Depending on the task, we set up automatic reruns for when the robot stops. Here, the robot reruns, skips reports already sent, and notifies your staff member in charge once sending is complete.

If the screen changes: When a screen change keeps the robot from continuing, it stops the run and sends a notification. Once you let us know, MedSync modifies the robot to match the new screen.

If the password has expired: An expired password means the robot can’t sign in, so it sends a notification without starting the run. We ask your staff member in charge to manage the dedicated robot account and its password.

How this diagram works

The processing, logs, and notifications for all five scenarios are written into the page, but only the scenario you select is shown. With animation stopped, you can still read the same content.

radio + :has() (works without scripts) / CSS animation (the status circles change color in turn) / View Transitions (switching scenarios)

In older browsers, the five scenarios are stacked vertically.

*This is a sample to illustrate the flow. The facility, times, and patients are fictional. How the robot behaves when it stops, and how it notifies staff, are designed to fit each task and workflow.

M-AUTOMATE

Pair RPA with AI to automate summaries and drafts

RPA is best at rule-based operations. For tasks such as reading records to judge whether they meet certain conditions, or summarizing text according to your facility’s rules, we add AI. M-Automate is the service that builds these AI steps into RPA workflows.

The basic M-Automate workflow

  1. 1By: RPARPA retrieves records

    Pulls the required fields and records from the EHR and other business systems.

  2. 2By: AIAI creates drafts

    M-Automate reads the records according to your facility’s rules, extracts and summarizes the information needed, and creates document drafts and lists.

  3. 3By: PersonA person reviews and edits

    A healthcare professional reviews the draft and any points that need case-by-case attention, edits as needed, and finalizes it.

  4. 4By: RPARPA files, exports, and sends

    Transfers and files the reviewed and edited text, then exports it to PDF, saves, sends, and prints it.

RPA gathers the information, and AI drafts the document according to your facility’s rules. After a person reviews and edits the draft, RPA takes over again to file, export, and send it.

This is the basic flow. For some tasks, the draft may be filed or exported to PDF before a person reviews it. Before anything is sent outside your facility, a person reviews and edits the content.

About AI drafts

What the AI creates is a draft. Physicians, nurses, and other healthcare professionals review and edit it and have the final say.

We adjust the wording to match each facility’s writing style.

*As of September 2026, M-Automate has been in operation at client medical institutions, and we’re happy to discuss adopting it. It runs inside RPA robots and can also be added to robots already in operation.

SUPPORT

From workflow analysis to operation and improvement after rollout

Before development, we show you how much time can be saved and the expected return on investment (ROI). After deployment, we check how long the work actually takes and verify the results.

What MedSync covers

  • Workflow analysis

    We ask about the steps involved in the task and your facility’s rules, then estimate how much time can be saved. We separate what will be automated from what people will check.

  • Development

    We design the robot around your facility’s procedures and write the specifications. The work is done by RPA engineers who know how medical institutions operate.

  • Deployment

    We test the robot with real data from your site and verify the results before going live. We show your staff member in charge how to operate it so your facility can keep it running.

  • Maintenance and improvement

    We run health checks, and if a robot stops, we find the cause and get it running again. Depending on the task, we also set up automatic reruns. We modify robots to keep up with screen changes in your systems, Windows updates, and changes to your procedures or document forms. We also handle problems with the robot PC. For ongoing operation, we have a privacy policy and a service level agreement (SLA) that defines the scope of support and service levels.

Who does what: your staff and MedSync

To keep the robots running reliably, your facility and MedSync each have a defined role. Notifications from the robots go to your staff member in charge. When a modification is needed, they contact us and we take care of it.

PersonWhat we ask your staff to handle

  • Running the robots day to day and checking the results
  • Following up on anything not sent or not processed
  • Managing the dedicated robot account and its password
  • Telling MedSync about EHR or other system updates and screen changes

MedSyncWhat MedSync handles

  • Analyzing workflows, then designing and developing robots
  • Investigating stops, restoring operation, and running health checks
  • Modifying robots for screen changes, Windows updates, and form changes
  • Answering questions about operating the robots

Support options

Choose from end-to-end support covering workflow analysis through operation and improvement, development-only projects, or online support for in-house development.

How patient information is handled

The robots run on PCs at your facility and operate the screens of your EHR and other systems. Run logs and lists are designed to hold only the patient information needed. For how data is handled when AI is added, see our information security page.

Development and operation focused on medical institutions

As an authorized reseller of the RPA tool BizRobo!, MedSync focuses its development and operation work on medical institutions. As of September 2026, at least 40 types of robots we developed in-house are at work in clinical settings.

FAQ

Frequently asked questions

We can hold meetings and consultations about operation online.

What kind of tasks do you start with?

We start with routine tasks done every day or every month. For hospitals, clinics, and home-visit nursing stations alike, we first identify the tasks to automate.

Do we need to replace our EHR?

We don’t assume a replacement. We suggest ways to make the most of the EHR and business systems you already use. Because RPA works through the screen, automation may be possible even without an API (a way for systems to exchange data). We’ll check your system and network environment and let you know whether it can be done.

Who handles daily operation?

We ask your staff member in charge to handle day-to-day runs and check the results. At deployment, we explain how to operate the robots and suggest a setup in which several people at your facility can run them. Troubleshooting and modifications for screen changes or procedure changes are covered under a maintenance and improvement contract.

If a robot stops, what happens to the work?

Your staff member in charge can tell that it stopped from the run log and, depending on the task, from a notification sent by the robot. Results are logged for each patient, so you can see which patients haven’t been processed yet. For tasks set up with automatic reruns, patients already processed are not processed twice. If a modification is needed, for example because a screen changed, MedSync handles it.

Could documents go to the wrong recipient?

For tasks that send documents by fax, the robot uses the recipient registered in the EHR and the fax number on file. If a patient has no recipient registered, the robot does not send that document. It adds the patient to a list and notifies your staff member in charge. We also design the robot so that documents already sent are not sent again.

Do we need a robot PC?

We recommend setting aside a dedicated PC to run the robot and keeping it on at all times. We’ll check your environment and advise on the PC and network settings needed for deployment.

Does patient information leave the facility?

When only BizRobo! robots are used, processing takes place on PCs at your facility. Patient information in the robot’s logs and lists is limited to what is necessary. When AI (M-Automate) is added, processing takes place in an environment dedicated to your facility, on a cloud in Japan. Neither your inputs nor the AI’s output is used to train AI models. For details, see our information security page.

Will this take work away from our staff?

Robots handle rule-based operations. Your staff continue to review content and make decisions, as they do now. MedSync helps you automate not to cut staff, but to raise the quality of care. We recommend deciding before adoption how you’ll use the time saved.

We already use RPA. Can you help?

Yes, we can help. You can also choose online support for in-house development. If you use BizRobo!, AI (M-Automate) can be added to robots already in operation.

When will we know the expected ROI?

We share it before development begins. We ask about the target tasks and how long they take now, then explain the expected ROI along with the estimated time savings.

CONTACT

Tell us which tasks eat up your time

Documents, tallying, or checks: if a task follows set steps, we’re glad to talk it through.

How a consultation works
  1. We ask which tasks you’d like to automate. Even one task is fine.
  2. We review your actual steps and systems and tell you if RPA can operate them.
  3. We estimate how much working time automation can save.