Australia · Working adults · Agents built around privacy

AI agents course for professionals

AI agents promise to read the inbox, draft the replies, update the system and file the report. For a working professional in Australia, the first question is not which tool is cleverest. It is what the agent will see, where that information goes, and who checks the result. Our AI agents course for professionals teaches agent building from the ground up, using automation that saves real hours, and it puts Australia's privacy expectations at the centre rather than in a footnote. This page summarises what the privacy regulator has published about AI products, tests a common safeguard on 500 invented customer messages, and sets out how the course is structured.

Book the free first class Book a priority demo See the experiment Ask on WhatsApp

Live teaching since 2020 · 10,000+ students · agents designed with the data question asked first

In short

An AI agents course for professionals teaches working adults to build agents that use tools to complete tasks, while handling the data they touch responsibly. In Australia, the privacy regulator's guidance says the Privacy Act applies to all uses of AI involving personal information, and recommends not entering personal information into publicly available generative AI tools. Modern Age Coders teaches this live online to Australian professionals, and the first class is free.

Start here

Three agent courses for working professionals

A coding-agents course for technical staff and two platform courses for organisations that already use those platforms. Each card opens the syllabus.

Codex & Claude Code course thumbnail

AGPR / 01

AI Coding Agents for Professionals

Directing coding agents well: specifications, review, tests and safe permissions. Adults only.

Open the syllabus →
Copilot Studio Agent Masterclass for Professionals course thumbnail

AGPR / 02

Copilot Studio Agent Masterclass

Agents inside a Microsoft workplace, taught one to one around your own process.

Open the syllabus →
Build AI Agents with Google Gemini Enterprise course thumbnail

AGPR / 03

Agents with Gemini Enterprise

Agents for Google Workspace organisations, taught one to one.

Open the syllabus →

Privacy duties

What the OAIC expects when AI meets personal information

The Office of the Australian Information Commissioner has published guidance on privacy and commercially available AI products. Anyone building agents at work should read it in full; these are the points that shape agent design most.

OAIC guidance, dated 21 October 2024, read 13 October 2026
The guidance saysWhat it means for an agent
"The Privacy Act applies to all uses of AI involving personal information."Every prompt and every tool result is in scope if it contains personal information
The OAIC recommends that organisations "do not enter personal information, and particularly sensitive information, into publicly available generative AI tools"Public chat tools are the wrong place for customer text
"AI products should not be used simply because they are available."Start with the task, not the tool; sometimes no AI is the answer
Due diligence "should not amount to a ‘set and forget’ approach"Agents need monitoring and periodic review after launch
Only "the minimum amount of personal information sufficient" for the purposeSend the agent fields, not whole records
Inferred or "artificially generated information" about an identifiable person "constitutes personal information"An agent's wrong guess about a customer is still their personal information

Accuracy is a duty, not a nice-to-have

Under APP 10, the guidance says, organisations "have an obligation to take reasonable steps to ensure the personal information collected, used and disclosed is accurate". An agent that drafts letters about customers inherits that obligation.

That is why every agent in our course has an evaluation set and a human review step before anything reaches a customer.

Tell people when they are talking to AI

The guidance also asks businesses to ensure that "public facing AI tools (such as chatbots) are clearly identified as such to external users such as customers."

We treat that as a design requirement: the agent says what it is, and a person can always be reached.

Source: Office of the Australian Information Commissioner, Guidance on privacy and the use of commercially available AI products, read 13 October 2026. This page summarises; it is not legal advice.

A worked example

Redaction caught every email and phone number, and none of the names

A common plan is to "strip the personal details" from text before an agent sends it to an AI service. We tested how well that works on 500 invented customer-support messages.

We wrote eight message templates and filled them with invented names, @example.com email addresses, mobile numbers, street addresses in a made-up suburb and order numbers: 500 messages containing 807 name mentions, 240 email addresses, 190 phone numbers and 199 addresses. No real person or customer appears. We then tried four ways of redacting them before they would be sent anywhere.

Personal details removed from 500 synthetic messages (our calculation)
Redaction methodNames removedEmails, phones, addressesOrdinary words wrongly removedMessages still containing a name
Patterns for emails, phones and addresses0 of 807all0500
Patterns plus a list of 50 common names389 of 807all0342
Patterns plus every name we used, any letter case807 of 807all3770
Send only the order number and issue typenothing to removenothing to remove00

Patterns are good at patterns

Email addresses, Australian mobile numbers and addresses with a state and postcode have a shape, and simple patterns caught every one. Names have no shape, so patterns caught 0 of 807 of them.

Name lists leak or overreach

A list of 50 common names still left a name in 342 of 500 messages. A complete list matched without regard to case removed 377 ordinary words, led by "will" (137), "hope", "may", "mark" and "bill", because they are also names.

The simple baseline wins

Sending only the fields the task needs, here an order number and an issue category, leaves nothing to redact. It is what the OAIC means by the minimum amount of personal information, and it beat every redactor we built.

Why the perfect result is a trap

Matching our full name list with exact capitals scored 807 of 807 with no wrong removals. That only happened because we invented the names from that list; real customers do not come with one.

A test that the method cannot fail is not a test. We show it so learners recognise the trap in vendor demonstrations, where the sample data has often been chosen to suit the tool being sold.

What professionals take away

Design the agent so personal details never need to leave your systems. Where free text must be processed, use an approved, contracted service rather than a public tool, and measure redaction before trusting it.

Count what slips through, on data that looks like yours, before any real data is involved. Then write the count down, because the same test should be run again whenever the agent, the model or the data changes.

Designing agents

Five design questions to answer before building an agent

Most failed workplace agents fail on the first question, not the last.

Questions we answer for every agent in the course
QuestionWhy it mattersA good answer looks like
Is an agent needed at all?A fixed process is better served by a script or a formA written comparison with the simplest option
What data will it see?Privacy duties attach to every input and outputA field list, with personal fields removed or justified
What can it change?Irreversible actions need a person to approve themRead-only by default; writes behind confirmation
How will we know it works?Accuracy is an obligation, not a hopeA test set with known answers, rerun on every change
Who reviews it later?Due diligence is not set and forgetA named owner and a review date

Students and younger learners have a separate AI agents course for students, which concentrates on tool-call checks and assessment rules.

The course

How the professionals course runs

Six stages, each ending in something usable at work, built on data that is either public or invented.

Course outline
StageWhat you buildWhat you learn to check
1. Task firstA written workflow and the simplest non-AI versionWhether an agent beats it
2. ToolsSmall functions the agent may call, read-only firstPermissions and logging
3. The agentA loop on a test model, then on an approved serviceEvery step replayable
4. Data handlingField-level minimisation and measured redactionWhat still leaks
5. EvaluationA task set with known answersAccuracy, cost of errors, drift over time
6. HandoverDocumentation, owner, review dateThat someone else can run it

Technical staff

Python, APIs and coding agents, with code review and tests as non-negotiable.

Business users

Platform agents in tools your organisation already licenses, with the same design questions.

Managers

Enough building to judge proposals: what to ask, what to measure and when to say no.

Honest limits

What this course is not

Clear boundaries protect learners and their employers.

Not legal advice

We summarise published guidance and teach engineering practice. Questions about your organisation's obligations belong with its privacy officer or lawyers.

We never ask learners to bring real customer or employee data into lessons.

No outcome promises

We do not promise promotions, job changes or savings figures. We promise working agents, measured honestly, that you understand well enough to defend.

Organisations must still make their own decisions about deployment.

Progress

Four levels of workplace agent skill

Most professionals reach the third level within one or two terms of weekly lessons.

From user to accountable builder
LevelWhat you can do
1. Informed userExplain what an agent sees, does and stores
2. BuilderBuild a read-only agent with logged tool calls
3. Responsible builderAdd minimisation, measured redaction, evaluation and approval steps
4. OwnerDocument, monitor and review an agent others depend on

Do I need to code?

For the technical track, basic Python helps; we teach it if needed. Platform courses need no coding to start.

Every track uses the same design questions.

Can you train our team?

Small groups from one organisation can be arranged. Lessons use public or invented data only.

Ask in the free class and we will propose a plan.

The catalogue

AI and automation courses for professionals

Adults only. Choose by the platform your work already uses, or by the depth you need.

I

Agents

Building and directing

AGPR / AGT / 01

AI Coding Agents for Professionals

Specify, review, test.

Open the syllabus
II

Foundations

Under the agent
III

Around the work

Data and teams

AGPR / WRK / 03

Data and AI for Non-Programmers

Analytics without code.

Open the syllabus

How lessons run

Lessons around the working day

Teaching happens from India in hours that match Australian evenings, weekends and some lunchtimes. We schedule in your state's time and keep lessons aligned through daylight saving.

Evening

After work, weekly.

Weekend

Longer build sessions.

Lunchtime

One to one, by arrangement.

Free first class

A small read-only agent and an honest view of fit.

Small groups

Professionals at a similar level.

One to one

Around one process or one platform.

Invented data

No real customer records in lessons.

Measured, not assumed

Every safeguard is tested.

People in charge

Approval steps before anything changes.

Student work

Work by our learners

Projects our students built and published. The student labs list many more.

NutriLife AI nutrition coach project screenshot

AI and ML

NutriLife

An AI nutrition coach that reads what you eat and works you toward a target.

by Bhavya · Open it

Misti AI chatbot for maths and coding screenshot

AI and ML

Misti

A chatbot that answers mathematics and programming questions, built and deployed by a student.

by Harshit · Open it

GuardianX AI internet safety assistant screenshot

AI and ML

GuardianX

An assistant that helps a young person recognise unsafe situations online.

by Vivaan · Open it

SkyCast weather forecast application screenshot

Web app

SkyCast

A weather forecasting site with live conditions for any location.

by Krish

Fees

Fees for the professionals agents course

Professionals in Australia pay monthly in US dollars. No upfront charge applies, and you may end the arrangement after any month.

Free first class

USD 0

no card required

  • A small read-only agent
  • A frank view of fit
  • No card required
Book it

Group batch

USD 100

a month, billed in US dollars

  • Small groups of professionals
  • Design reviews between learners
  • A regular teacher
  • End after any month
Start here

One to one

USD 150

a month, billed in US dollars

  • Built around one workflow
  • Platform-specific if needed
  • Evening, weekend or lunchtime
Enquire

What families say

Rated 4.9 across 547 Google reviews

What learners and parents wrote on Google, quoted without changes.

★★★★★

"The one step solution for my son. Modern Age Coders make learning coding so simple that kids love it. The teachers explain complex concepts clearly with practical exercises and interactive content."

Ria Mukherjee

Parent

★★★★★

"Modern Age Coders has been a game-changer for me. I struggled to grasp IT concepts and coding before joining, but their classes transformed everything. I can now confidently write complex programs with ease."

Samriddha Mondal

Student

★★★★★

"One of the most wonderful education centres out there. Education is not limited to school syllabus but focuses on skill development."

Vansh Agarwal

Student

★★★★★

"My child Dhairya is really enjoying the Modern Age Coders classes. This is his first online class and he eagerly looks forward to it. I can already see his improvement, and the teachers are very cooperative."

Sonam Oswal

Parent of Dhairya

★★★★★

"Modern Age Coders have wonderful teachers who teach in a clear, easy and practical way. The teacher boosts students' confidence and inspires them to learn without hesitation."

Sonu Goyal

Parent

★★★★★

"I highly recommend this computer coding class! The teachers are incredibly knowledgeable and passionate about coding."

Ritu Kedia

Parent

Professionals ask

Questions about AI agents courses for professionals in Australia

What is an AI agents course for professionals?

It is a course that teaches working adults to design, build, evaluate and maintain AI agents, programs that use AI to choose and run tools, for real workplace tasks, with privacy and human review built in.

Can I put customer information into an AI chatbot?

The OAIC recommends that organisations do not enter personal information, and particularly sensitive information, into publicly available generative AI tools. Your organisation's own policies also apply.

Does redaction make it safe to send text to AI?

Not reliably. In our test on 500 invented messages, patterns removed every email, phone number and address but no names. Sending only the fields a task needs worked better.

Does the Privacy Act apply to AI?

The OAIC's guidance states that the Privacy Act applies to all uses of AI involving personal information, including information an AI system generates or infers about an identifiable person.

Do I need programming skills?

Not for the platform tracks. The technical track uses Python, which we can teach from the start.

How was the redaction test built?

We wrote eight message templates and filled them with invented names, example.com email addresses, mobile numbers in Australian formats and addresses in a made-up suburb, then measured what each redaction method removed and what it missed.

Is this legal advice?

No. We summarise published guidance and teach engineering practice. Your organisation's privacy officer or lawyers advise on obligations.

Which platforms do you cover?

Python-built agents, coding agents, and one-to-one courses for Microsoft Copilot Studio and Google Gemini Enterprise for organisations already using them.

What does it cost?

The first class is free with no card. Monthly fees are listed in the fees section, for groups and one-to-one.

Also on this site

Related pages

Related AI and adult learning pages.

Coding classes in Australia

The national index of our pages.

AI agents course for students

Tool-call checks and assessment rules.

Coding classes for adults

Python, data and the web around work.

Learn to think with AI

Judgement before automation.

AI and machine learning in Australia

Building models, not just agents.

One-on-one coding tutor

Private lessons around one workflow.

Start here

Book a free first class

Leave a number and a time. We build a small read-only agent together in the free lesson and talk honestly about whether the course fits your work.

To learn more first, read about our teaching or look over the courses.

WhatsApp us · +91 91233 66161 · connect@modernagecoders.com

Our people are in India, so you will see a +91 number, and there is no office in Australia. WhatsApp from an Australian phone normally brings the fastest response.

Free, no card, no obligation.

Want a class sooner? Book a priority demo

A priority demo is a full class of about 45 to 60 minutes. After the class the mentor sends a written skill report and a personal learning roadmap on WhatsApp. If you enrol, the demo fee is adjusted against your first month's fee. Share your demo Order ID with us on WhatsApp and we take it off.

Book a priority demo

Keep exploring Modern Age Coders

Related courses and guides

Learn more

Free resources

From the blog

Start here

Free classPriority demoCall us