MLNew
Ready to build models yourself?
The Practical AI track: 11 lessons where you write machine-learning models in Python that run in your browser, then AI inside ERP and Power BI.
Start the track
AI · Start here
AI, simply explained
What is AI, really? How does a machine “learn”? And how can you use it as an MIS student, in your studies and your career? A short explanation with no equations and two hands-on demos.
NewThe prompt-writing guide
Open the guide
Why the way you write changes the answer, the six parts of a strong prompt, a builder for your own prompts, and templates for SQL, requirements and studying.
NewThe prompt library
Open the library
30+ ready study prompts: understanding lessons, revision, programming, systems analysis, academic writing and interviews. All of them make AI teach you instead of doing the work for you.
So what is AI?
Artificial intelligence (AI) is any system that does a task we used to need human intelligence for: understanding language, recognising an image, forecasting a number or writing text. The terms you hear every day are circles inside one another:
- AI · Artificial intelligence
The whole field: any system that imitates an intelligent ability, even hand-written rules.
- ML · Machine learning
Instead of writing the rules ourselves, the system works them out from data.
- DL · Deep learning
Machine learning with large neural networks, strong on images, audio and text.
- GenAI · Generative AI
Models that generate new content: text, images, code. ChatGPT, Claude and Gemini are examples.
How does a machine learn? Try it
We have data on 16 students: hours studied and whether they passed. We want a rule that predicts passing from study hours. Move the slider and see how many students you predict correctly, or let the machine try every value and pick the best.
Passed✕FailedWrong prediction
Rule: studied 3 hours or more → passes
Accuracy75%12 of 16 predictions correct
That's machine learning in a nutshell: data + a goal + trial. The machine tried many rules and kept the most accurate one. Real models use hundreds of features and millions of examples, but the idea is the same. Notice that even the best rule gets some students wrong: no model is 100% accurate.
How do ChatGPT and Claude work?
Large language models (LLMs) were trained on a huge amount of text and learned one thing extremely well: predicting the next word. When you ask a question, they pick a likely word, then the next, and so on until the answer is complete. Try building a sentence:
A database▍
Tap a word to add it. The numbers are illustrative.
- They don't “know” like people do: they generate text that resembles what they were trained on, so they can state something wrong with full confidence. This is called hallucination.
- Their knowledge has a cut-off: they were trained on data up to a certain date, unless they're connected to search.
- Better questions, better answers: the clearer your request and the more context it gives, the better the result.
Where AI shows up in companies
As an MIS student you'll usually meet AI as part of the systems you analyse and implement, not as something separate:
Forecasting
Forecasting sales and inventory demand in ERP and supply-chain systems.
Fraud
Banks flag suspicious transactions in real time from spending patterns.
Recommend
“Customers who bought this also bought…” in online stores.
Chatbots
Assistants answer common questions and hand hard ones to a person.
Documents
Pulling data from invoices and contracts straight into the system.
Security
Spotting unusual network behaviour before it becomes a breach.
Use it in your studies (the right way)
- Explain it to me: ask for a hard concept with real-world examples, such as normalisation or many-to-many relationships.
- Quiz me: have it write questions on a topic before an exam, and answer them yourself first.
- Why doesn't my code work? Paste the SQL or Python error and ask for the reason, not just the fix.
- A first draft: user stories from a system description, or entities for an ERD, then check them yourself in the ERD Studio.
A vague request versus a clear one
Vague
Write me SQL for sales
Clear
I have two MySQL tables: orders(order_id, customer_id, order_date) and order_items(order_id, product_id, quantity, unit_price). Write a query that returns total sales per month in 2024, newest first, and briefly explain each part so I understand it.
The clear request gives context (tables and columns), the exact task (per month, 2024, ordered) and the answer format (with an explanation).
NewThe prompt-writing guide
Open the guide
Why the way you write changes the answer, the six parts of a strong prompt, a builder for your own prompts, and templates for SQL, requirements and studying.
NewThe prompt library
Open the library 30+ ready study prompts: understanding lessons, revision, programming, systems analysis, academic writing and interviews. All of them make AI teach you instead of doing the work for you.
AI in MIS careers
AI isn't replacing MIS roles, but it is changing how the work gets done. People who use it well become faster and more valuable:
| Junior Business Analyst | Summarises stakeholder interviews into draft requirements and user stories. |
|---|---|
| Junior Data Analyst | Writes queries and analysis code faster and uses AI features inside Power BI and Excel. |
| Junior ERP Consultant | Implements AI built into SAP, Oracle and Odoo, such as forecasting and invoice reading. |
| SOC Analyst (Tier 1) | Uses AI threat-detection tools and protects systems from attacks that use AI. |
| Junior IT Auditor | Writes company AI-use policies and assesses their risk and compliance. |
Watch out for these risks
Wrong information
Check any number, reference or code before relying on it. Run the code and test it.
Privacy
Don't paste customer, employee or company-confidential data into a public tool. Follow your organisation's policy and data-protection law.
Bias
Models learn from biased data, so they can be unfair in decisions such as hiring or loans.
Academic integrity
Know your university's policy. Use it to understand and review, not to hand in work for you.
Check yourself
1. What's the key difference between machine learning and regular programming?
2. ChatGPT and Claude are examples of:
3. Why can a language model state something wrong with confidence?
4. In the study-hours demo, what did the machine do to “learn”?
5. Which is the best use of AI for a university assignment?
Starter AI certifications
Three foundational certifications that don't need advanced programming and suit MIS students. Details are in the certification guide.