AI for Healthcare Staff is a hands-on course in using ChatGPT, Gemini, Copilot and NotebookLM for non-clinical hospital work such as letters, meeting minutes, appointments, duty rosters and patient information drafts, while masking health data under PDPA and keeping staff review on every output. It suits administrative, medical records, nursing and support staff in hospitals and clinics.
Hospital staff spend a great deal of time on work that is not direct patient care: letters, memos, meeting minutes, appointments, duty rosters, quality documents and answering questions from service users. Generative AI can reduce the time this paperwork takes, but health information is highly sensitive personal data and AI can produce wrong information. Using AI in a hospital therefore needs clear boundaries, no patient-identifiable data in AI tools, and staff review of every piece of work before it is used.
This course covers non-clinical work only and does not cover diagnosis or treatment advice. Learners use ChatGPT, Gemini, Copilot and NotebookLM with sample documents, from masking data in line with PDPA and writing prompts, to drafting letters and meeting minutes, drafting patient information for the relevant staff to review, handling appointments and rosters, and searching manuals and quality documents. The day closes with an AI plan for your unit. (1 day, 6 hours, Beginner level.)
What you’ll gain
Separate tasks AI can help with from clinical work where AI must not make decisions
Mask patient-identifiable data before using AI and explain PDPA principles for health data
Write prompts for paperwork, administration and communication in a hospital
Use AI to draft letters, memos, meeting minutes and appointment messages
Draft easy-to-read patient information and send it to the relevant staff for review before use
Search and summarise manuals, regulations and quality documents with NotebookLM and check the sources
Plan how to bring AI into your unit with human checkpoints
Who this course is for
Administrative, medical records, finance and appointment centre staff in hospitals
Nurses and support staff who handle paperwork, duty rosters and coordination
Quality improvement, human resources and public relations teams in healthcare facilities
Supervisors who want to set safe AI practices in their units
Prerequisites
Everyday use of a computer, web browser and office software
No programming or AI background is needed
An account for at least one of ChatGPT, Gemini or Microsoft Copilot (a free plan is enough)
A Google account for NotebookLM
Curriculum
Course Details
A 1-day, 6-hour hands-on course (09:00-16:00) at Beginner level for hospital and healthcare facility staff. It covers non-clinical work only, with no diagnosis or treatment advice, and differs from the clinic courses that focus on marketing and business. Learners use ChatGPT, Gemini, Microsoft Copilot and NotebookLM with their own or organisation-approved accounts, and every lab uses sample documents with no real patient data. Learners take home the course handbook, a prompt library, a checklist for masking data and reviewing AI drafts, and a template for an AI plan for their unit.
Day 1: Using AI Safely for Hospital Paperwork
Section 1AI in the Hospital: Where It Helps and Where It Must Not Be Used
Where AI helps: paperwork, administration, communication, appointments and draft patient information
A clear line: no AI for diagnosis, treatment orders or medical advice
Hallucination and the impact of wrong information in a health setting
Every AI draft must be checked by the responsible person before use
Examples of time-consuming support tasks that are good places to start
Section 2Lab: Health Data, PDPA and Masking
Health data is highly sensitive personal data that needs extra protection
Never enter names, HNs, ID numbers, test results or anything identifying a patient into public AI tools
Personal versus organisation accounts, and the tools your facility allows
Lab: mask a sample document so no individual can be identified before using AI
Section 3Lab: Prompting for Hospital Work
Prompt structure: role, context, task, output format and constraints
Set the right register for internal documents and for communicating with service users
Add examples and ask the AI to ask back when information is missing
Lab: compare ChatGPT, Gemini and Copilot on the same task
Section 4Lab: Letters, Memos and Meeting Minutes
Draft memos, coordination letters and internal notices from short briefs
Summarise meeting notes into resolutions, owners and deadlines
Double-check names, positions, figures and dates that AI may invent
Lab: turn a sample unit meeting into minutes using a standard form
Section 5Lab: Drafting Information for Patients and Families
Rewrite content already approved by the care team so it is short, clear and friendly
Adjust language for older readers and draft English versions for international service users
All medical content must be reviewed and approved by the relevant staff before release
Lab: draft a leaflet on preparing for an appointment from sample information
Section 6Lab: Appointments, Duty Rosters and Statistics
Draft standard messages for appointments, rescheduling and preparation
Use AI to help draft a duty roster from sample rules and find conflicts
Summarise appointment statistics from de-identified tables in Excel or Google Sheets
Lab: analyse a sample appointment table and write a short report
Section 7Lab: Manuals and Quality Documents with NotebookLM
Build a notebook from operating manuals, regulations and quality documents with no patient data
Ask questions with cited sources and always open the original
Summarise into FAQs, checklists and orientation material for new staff
Share notebooks only with the people involved and review access regularly
Lab: build a unit knowledge base from sample documents
Section 8Lab: Service Requests and Complaints
Draft FAQ answers about procedures, entitlements and how to get in touch
Categorise masked complaints and draft polite replies
Keep an empathetic tone and never promise beyond your authority
An authorised person reviews and approves before any reply is sent
Lab: build a set of standard replies for a sample service
Section 9Lab: Internal Reports and Presentations
Summarise performance reports from de-identified data so executives can read them easily
Build a slide outline with Copilot or Gemini from the summarised report
Draft concise newsletters and internal notices for the unit
Check every figure, unit name and statement before presenting
Lab: make a monthly results slide deck from sample data
Section 10Workshop: An AI Plan for Your Unit
Choose 2-3 frequent, low-risk paperwork tasks
Design a new workflow that marks where data is masked and where people review
Workshop: write the new workflow for your chosen tasks with the prompts you will use
Set measures: time saved, work quality and data accuracy
Summarise your unit's Do and Don't list and present the plan for feedback
Schedule & training options
For individuals — public rounds
No public rounds are open right now. Join the waiting list and we will contact you first when the next round opens, or ask us on LINE. Or call 02-570-8449 or 088-807-9770
Which hospital tasks can AI help with, and which should it not be used for?
AI suits non-clinical work such as drafting letters and memos, summarising meeting minutes, writing appointment messages, organising duty rosters and drafting easy-to-read patient information, always with staff review. Diagnosis, treatment choices and individual medical advice should not be decided by AI, and data that identifies patients must never be entered into general AI tools.
How do you mask patient data before using AI, in line with PDPA?
Health data is sensitive personal data under PDPA, so anything that identifies a patient must be removed or replaced before it goes into an AI tool, including names, hospital numbers, national ID numbers, dates of birth, addresses, phone numbers and any detail that reveals who they are. Use placeholder codes instead, double-check with a checklist before sending, and use only tools or accounts your organisation has approved.
How do you use AI to draft easy-to-read information for patients?
Start with accurate source content from guidelines or documents your unit has approved, and state the audience, such as older patients or family carers. Ask the AI to use short sentences, avoid or explain medical terms and organise the content into headings or steps people can follow. Then have doctors, nurses or other relevant staff check it for accuracy before every use.
Who is the AI for Healthcare Staff course for, and what should I know beforehand?
It suits hospital administrative, medical records, finance and appointment centre staff, nurses and support staff who handle documents and duty rosters, quality, HR and communications teams, and supervisors who need to set safe AI guidelines. It is a beginner course: general computer and office skills are enough, with no programming or AI knowledge required.
What will I be able to do after the AI for Healthcare Staff course?
You will be able to separate tasks AI can help with from clinical work, mask patient data before using AI, draft letters, memos, minutes and appointment messages, draft patient information for review, search and summarise manuals and quality documents with NotebookLM, and plan AI use in your unit with human checkpoints. Participants receive a Certificate of Completion with its own number that can be verified online.
Can the AI for Healthcare Staff course be run for our own hospital team?
Yes, it can be run in-house for a hospital or clinic, with examples adapted to each unit's work, such as medical records, the appointment centre or quality improvement. All labs use mock documents with no real patient data. Request a quotation from this page, or contact LINE @itgenius or call 02-570-8449.
Public sector work involves a large volume of documents that are produced again and again: official letters, internal memos, meeting minutes, performance reports and replies to citizens. Generative AI tools such as ChatGPT, Gemini, Copilot and NotebookLM can…
Thailand's Personal Data Protection Act B.E. 2562 (PDPA) is now fully in force, and most of the day-to-day workload falls on IT teams and data custodians: knowing what personal data the organization holds and where, maintaining records of processing…
ChatGPT has become a tool many people open every working day, yet most still use it one question at a time. Today it can read files, analyze data, produce Word, Excel and PowerPoint files, connect to email and calendars, carry out in-depth research and run…