awsfor Healthcare
register
Scan to register interest

Scan to register · opens full-screen 16:9

📊
Quick Sight Dashboard
Equipment Operations — uptime, PM, repair cost
● LIVE
🧩
Use case A — Knowledge
RAG over device manuals
● Open Space A
📋
Use case B — Clinical Docs
Summarize / extract clinical docs
● Open Space B
⚡
Quick Flow — Referral Intake
extract + triage + draft email
● Open Flow
🔄
Quick Automation — PM Alert
weekly maintenance reminder
✎ paste Automation URL
🏥
Quick Flow — OPD Registration
staff types name, search history + triage
● Open Flow
📱
Quick App — HCLS Intelligence Hub
custom healthcare intelligence app
● Open App
🩺
Quick App — Patient 360°
whole patient on one screen + AI watch
● Open App
🎭
Quick App — Hospital Command Center
equipment cockpit + AI fix/fund priorities
● Open App
⚙️
Admin (back-office)
how it works · setup
presenter guide

Demo Script — questions live inside each use case

1 Use case A — Medical Knowledge Assistant Space + Quick Index

Q&A over device manuals / SOPs / error codes with source citations (RAG)

How to show: open Space "Biomedical Knowledge Assistant" → type a question → point to the citation

Sample questions (click Copy)

What does error E-305 on the Acme V-100 ventilator mean, and how do I fix it?
At what values must the Acme V-100 flow be calibrated, and what are the pass criteria?
What daily checks are required for the Northwind DF-300 defibrillator?
What is the difference between infusion pump errors P-21 and P-22?
How do I fix error U-301 on the Globex US-700 ultrasound?
How often must high-risk equipment have PM, and which devices are they?

2 Use case B — Clinical Document Intelligence Space + Agent

Summarize / extract / Q&A over clinical documents (referral / lab / discharge) in seconds

How to show (wow): keep referral-letter-002.pdf to drag in live and summarize instantly

Sample questions (click Copy)

Summarize this referral letter, including the reason for referral and current medications.
Which lab values are abnormal for this patient, and what do they indicate clinically?
Summarize the discharge summary as bullets: diagnosis, procedure, discharge meds, follow-up.
From the medication list, which meds changed at discharge and what cautions apply?
Compare patients referral-001 and referral-003 in terms of urgency.
Expected: catches abnormal values (high FBS/HbA1c/LDL/Troponin) + flags 'do not stop dual antiplatelet'

3 Quick Sight — ask data in natural language Data → Analyses → Dashboards

Executives ask in plain language → instant charts, no waiting on IT

How to show: open the Dashboard (LIVE button) or use the NL Q&A bar

Sample questions (click Copy)

Which hospital has the lowest average uptime?
How many devices are still Under Repair, and where?
Total repair parts cost this year by hospital.
Which month and hospital had the biggest uptime drop?
Which error code occurs most often?

4 Quick Flows — Referral Intake Assistant Flows

From one referral letter → 3 tasks in one click (reusable for every referral)

Test with referral-letter-001.pdf

  1. Input: referral letter (PDF)
  2. Extract patient data
  3. Triage urgency (urgent / non-urgent)
  4. Draft acknowledgement email + appointment

4B Quick Flows #2 — Patient Intake & Pre-visit Briefing Flows · เวชระเบียน

At registration, triage the patient, route them to the right staff, and auto-summarize their history for the doctor — so the exam starts faster.

How to show: upload patient-emr-somchai-001.pdf into Space B first, then run the flow with registration-intake-001.pdf as input.

  1. Triage: urgency level + recommend department & staff
  2. Retrieve & summarize the patient's history for the doctor
  3. Link today's symptoms to history, suggest what to check

Flow prompt — paste into Create flow

สร้าง workflow สำหรับจุดเวชระเบียน OPD ของโรงพยาบาล (โหมดเจ้าหน้าที่กรอกเอง)
ให้เจ้าหน้าที่กรอก: ชื่อ-นามสกุลผู้ป่วย (หรือ HN) + อาการที่มาพบวันนี้ + คำตอบซักประวัติเบื้องต้น + สัญญาณชีพ
ขั้นตอน:
1) ค้นหาผู้ป่วยจากชื่อ/HN ในเวชระเบียน (ประวัติการรักษาที่จัดทำดัชนีไว้) แล้วดึงประวัติที่ตรงกัน — หากไม่พบ ให้แจ้งว่าไม่พบประวัติ และ triage จากอาการที่กรอกเท่านั้น
2) คัดกรอง (Triage): จัดระดับความเร่งด่วน (ฉุกเฉิน/เร่งด่วน/ทั่วไป) + แนะนำแผนก/บุคลากรที่เหมาะสม พร้อมเหตุผลจากอาการ
3) สรุปประวัติให้แพทย์ (Pre-visit Briefing): โรคประจำตัว, ยาปัจจุบัน + ข้อควรระวัง/ยาห้ามหยุด, การแพ้ยา, หัตถการ/การรักษาล่าสุด, ผลแล็บสำคัญ, นัดติดตาม
4) เชื่อมโยงอาการวันนี้กับประวัติ เสนอสิ่งที่แพทย์ควรซัก/ตรวจเพิ่ม
Output: การ์ด routing (แผนก + ความเร่งด่วน) + สรุปประวัติ 1 หน้าสำหรับแพทย์
Create a workflow for a hospital OPD registration desk (staff-entry mode).
Staff enters: patient full name (or HN) + today's chief complaint + initial triage answers + vital signs
Steps:
1) Search the patient by name/HN in the indexed medical records and retrieve the matching history — if not found, say no history found and triage from the entered symptoms only
2) Triage: classify urgency (emergency/urgent/routine) + recommend the right department/staff, with reasoning from the symptoms
3) Pre-visit Briefing for the doctor: chronic conditions, current meds + cautions/do-not-stop, allergies, recent procedures, key labs, follow-up
4) Link today's symptoms to the history and suggest what the doctor should ask/examine
Output: a routing card (department + urgency) and a one-page history summary for the doctor
Expected: routing = Cardiology / Urgent (exertional chest pain + recent PCI); briefing flags dual antiplatelet must not stop, suggests repeat ECG/troponin (possible in-stent restenosis).

5 Quick Automation — PM Due Alert More → Automations

Recurring work done automatically — no one manually building reports

In the demo, click Run now to show results instantly

  1. data source = pm-schedule-upcoming.csv
  2. filter days_until_due ≤ 30
  3. summarize as a table
  4. Schedule Monday 08:00 → email the supervisor