College and university programmes
Learn beside your degree, or take the whole journey
Our college levels add the hands-on, interactive practice that lecture-and-lab courses leave out. We do not replace your university course. We help you understand it better, score higher and become employable.
Registration open · self-paced online
Basics / Fundamentals of Data Science, Machine Learning and AI
33 chapters, about 116 hours, with lessons, labs, quizzes, assignments, a project and an employability report. Study at your own pace with six months of access.
Four levels and a complete diploma
Hours are learning hours, including labs. Each level can be taken on its own.
| Programme | Hours | Who it is for | By the end, the learner can |
|---|---|---|---|
| Start here (S01-S06) | 19 h | Anyone new to data or computers | Use a spreadsheet to answer simple questions, read a flowchart and run a first notebook |
| Basic (B01-B26 + project) | 97 h | Second-year students; beginners from any field | Write small programs, query a database, summarise and chart data, build a first machine learning model, and score better in second-year exams and labs |
| Medium (M01-M19 + project) | 105 h | Third-year students; Basic completers | Train, evaluate and explain machine learning and deep learning models, including NLP, vision, reinforcement learning and big data, and finish an ML/DL project |
| Advanced (A01-A17) | 100.5 h | Final-year and master's students; Medium completers | Build, deploy and audit an AI system with applied and generative AI, BI and research methods, and present a decision-ready capstone and job portfolio |
| Advanced Diploma in Data Science, Machine Learning and AI | about 207 h + 8 h career | Learners who want the whole journey in one programme | Work job-ready across analytics, machine learning, deep learning and generative AI, with five milestone projects and a capstone |
Advanced Diploma · module by module
41 modules in five blocks and a capstone
Modules 01 to 07 use guided labs with the code given, Excel workbooks and SQL; Python starts in Module 08. A capstone milestone on the case company's data closes each block. Open a block to see every module.
AData analytics foundations
| Module | Title | Hours | Learners will be able to | Hands-on work |
|---|---|---|---|---|
| 01 | Introduction to Data Analytics | 3 h | Name the four types of analytics, follow the lifecycle, build a KPI tree, use AI assistants safely | Guided lab: a first look at the company; 30-question assignment |
| 02 | Data Collection and Storage | 3 h | Judge sources and bias, read grain and keys, pick formats and storage, mask personal data | Guided lab: where does the data live? |
| 02B | SQL for Data Analytics | 7 h | Join without losing rows, write CTEs and window functions, audit AI-written SQL | Two DuckDB labs on retail data |
| 02B+ | PostgreSQL and MySQL | 3 h | Run a server database: constraints, staged loads, indexes, roles, transactions | PostgreSQL lab in Colab; 100-mark graded assignment |
| 03 | Data Cleaning | 4 h | Measure data quality, fix labels, dates and duplicates by rule, handle missing values | Fix a messy orders export to the exact truth |
| 04 | Exploratory Data Analysis | 4 h | Describe variables, compare groups with margins of error, spot confounding, use automated EDA | Guided lab: what drives returns? |
| 05 | Tools of the Trade | 4 h | Map tools to the lifecycle; answer one question in a spreadsheet, SQL, pandas and Polars | One question, four tools, with timings |
| 06 | Fundamentals of Excel | 2.75 h | Write formulas that copy, use lookups and conditions, format and clean data | Regional store workbook |
| 07 | Advanced Excel | 2.75 h | Build pivot reports with slicers, honest charts, what-if analysis and Power Query | Store manager's pivot report |
| 08 | Python Basics | 4.5 h | Work in a notebook; use types, money arithmetic, strings and conditions; read errors | Lab: one order, three systems |
| 09 | Loops, Functions and Errors | 6 h | Loop over files, write and test functions, handle errors | Clean 18,516 export lines in pure Python |
| 10 | Lists and Tuples | 4.5 h | Build, slice, sort and search lists; know when a set is faster | Lab: what is in the basket? |
| 11 | Dictionaries and Sets | 3.5 h | Count, group and join with dicts; compare customer groups with sets | Lab: who buys, and where? |
| 12 | NumPy | 3.5 h | Compute without loops, filter with masks, work in 2-D, broadcast | Lab: every line, every week |
| 13 | Data Manipulation with pandas | 3.5 h | Select, clean, group, reshape and merge with checks; work with time | Build the customer 360 (8,000 rows) |
BVisualisation
| Module | Title | Hours | Learners will be able to | Hands-on work |
|---|---|---|---|---|
| 24A | Storytelling with Data | 6 h | Find the big idea, choose and declutter charts, keep axes honest, tell it on one page | Chart makeovers; one page for the CEO |
| 24B | Visualisation in Python | 8 h | Build charts in Matplotlib, seaborn and Plotly, styled once and reused | Weekly trading pack; interactive sales explorer |
| 24 | Power BI | 5.5 h | Shape data in Power Query, model it, write DAX measures, publish a dashboard | Regional sales dashboard |
| 25 | Tableau | 4.5 h | Use calculated fields, level-of-detail expressions, actions and story points | The same dashboard in Tableau, compared |
CMachine learning
| Module | Title | Hours | Learners will be able to | Hands-on work |
|---|---|---|---|---|
| 13A | Intro to ML and Regression | 4 h | Frame a supervised problem, set a baseline, fit and read a line, measure error honestly | Your first model, from footfall to bags sold |
| 13B | Multiple Regression | 8 h | Read effects holding others fixed, judge uncertainty, estimate price elasticity | Price and promotion effects; from model to pricing decision |
| 14 | Logistic Regression | 8 h | Read odds ratios, evaluate beyond accuracy, choose a threshold from costs | Who will stop buying?; filling in missing loyalty tiers |
| 15 | Trees and Ensembles | 8 h | Read trees, prevent overfitting, use random forests, XGBoost and LightGBM | Trees you can read; forests and boosting |
| 16 | Model Evaluation | 8 h | Cross-validate properly, test out of time, calibrate scores, explain with SHAP | Can we trust the score?; SHAP reason codes |
| 17 | Unsupervised Learning | 8 h | Segment with k-means and hierarchical methods, find outliers, reduce with PCA | Customer segments; PCA and complaint themes |
| 18 | SVM and KNN | 8 h | Use margins and kernels, choose k and distances, build a recommender | SVMs for churn and tickets; KNN recommendations |
| 19 | Time Series Forecasting | 8 h | Use ARIMA, SARIMA, Holt-Winters, Prophet, ML and Chronos-2; backtest; plan capacity | Forecasting the Mega Sale; Prophet, ML and foundation models |
DDeep learning, NLP and vision
| Module | Title | Hours | Learners will be able to | Hands-on work |
|---|---|---|---|---|
| 20 | Intro to Deep Learning | 8 h | Train networks in PyTorch; tune optimisers; fight overfitting; build CNNs | Neural-net churn model; a CNN that tags product photos |
| 21 | Deep Learning Architectures | 8 h | Use RNN, LSTM and GRU; explain vanishing gradients; build attention | Sequences of customer visits; reviews, LSTMs to attention |
| 22 | Natural Language Processing | 4 h | Represent text as numbers; build and evaluate text classifiers | Review sentiment; ticket routing and themes |
| 23 | Advanced NLP | 3.5 h | Fine-tune BERT-style models; extract entities from tickets | A BERT ticket router; entity extraction |
| 23A | Computer Vision | 3.5 h | Use transfer learning, ViT and CLIP; detect objects with YOLO | Tagging product photos with pretrained models |
EGenerative and applied AI
| Module | Title | Hours | Learners will be able to | Hands-on work |
|---|---|---|---|---|
| 26 | GenAI, Transformers and LLMs | 2 h | Explain how transformers and LLMs work; use Colab and Hugging Face open models | Tokens, next-token prediction, prompting |
| 26A | Generative Models | 4 h | Explain VAEs, GANs and diffusion; generate and judge synthetic images | Synthetic product images |
| 27 | Fine-tuning LLMs | 5.5 h | Decide when to fine-tune; run supervised fine-tuning with LoRA | Recipe assistant (LoRA); dialogue summaries (Flan-T5) |
| 28 | Advanced Fine-tuning and Evaluation | 7 h | Explain RLHF and DPO, control decoding, evaluate LLM outputs | Decoding and evaluation; voice assistant; DPO |
| 28A | RAG, Agents and a Chatbot | 8 h | Build RAG from scratch, measure retrieval, run a guarded agent loop | RAG for customer care; support agent and Gradio chatbot |
| 28B | Deployment and LLMOps | 4 h | Serve a model as an API, test and package it, watch drift, run LLMOps | Ship and monitor a model; LLMOps for the assistant |
| 28C | Responsible AI | 2 h | Check privacy, bias and fairness; document and govern AI systems | Audit a model and an assistant |
| 28D | Reinforcement Learning (optional) | 2 h | Explain rewards, policies and bandits; link them to RLHF | Dynamic pricing introduction |
29Capstone and career
| Milestone | What | When | What learners build |
|---|---|---|---|
| 1 | Data audit | After Block A | Audit the data: quality, keys, cleaning rules and a data dictionary |
| 2 | Dashboard | After Block B | A regional sales dashboard and a one-page story |
| 3 | Churn and demand models | After Block C | Predictive models that end in a business decision |
| 4 | Ticket triage | After Block D | Route and tag support tickets automatically |
| 5 | Support assistant | After Block E | A grounded assistant, deployed, monitored and audited |
| 29 | Capstone | End | Join the milestones into one end-to-end project and present it |
How learning is checked
- In class: lab questions with expected answers given, and "your turn" cells that change one thing
- After class: homework exercises and an assignment workbook of 20 to 30 questions
- Every module: a check-your-understanding quiz and key takeaways
- Across the course: five capstone milestones, one after each block, then the final capstone
- For trainers: an answer key with every expected number, so marking is the same across batches and institutes
Three ways to learn the same chapter
College Basic, Medium and Advanced chapters come in three coverage levels, so learners with different backgrounds can study together:
- No code: spreadsheets, visual tools and AI assistants
- Try a little code: fill-in-the-blank notebooks with hints
- Full code: write the Python yourself
Pick your path
| You are | Start with | Then | Main goal |
|---|---|---|---|
| Second-year engineering student in AI and data science, computer, IT or AI-ML | Basic | Medium in third year | Deeper understanding, higher grades |
| Third-year engineering student | Medium | Advanced in final year | Higher grades, ML/DL projects |
| Final-year engineering or master's student | Advanced | Career module, case studies | Strong project, placement |
| Student or graduate of another field (BSc, BCom, BBA, other engineering) | Start here + Basic in no-code mode | Medium, Advanced | Employability in data and AI |
| Learner wanting one complete programme | Advanced Diploma | Case studies | Job-ready data scientist or ML engineer |
Current tools, free to run
Labs run on free Google Colab with open models. Paid APIs appear only as optional extras that are switched off.
- Excel and Power Query
- SQL, DuckDB, PostgreSQL, MySQL
- Python, pandas 3, Polars, NumPy
- Matplotlib, seaborn, Plotly
- Power BI, Tableau
- scikit-learn, XGBoost, LightGBM, SHAP
- Prophet, Chronos-2
- PyTorch
- Hugging Face, BERT, ViT, CLIP, YOLO11
- Qwen and other open LLMs, LoRA, DPO
- RAG, agents, MCP, Gradio
- Git and GitHub, Ollama
Setup tracks: every computer ready before day one
Each track has a deck with trainer notes, a Windows, macOS and Linux guide with troubleshooting, and a setup check script that says pass or fail.
| Track | Time | What learners set up |
|---|---|---|
| ST01 Computer Ready | 1 h | Admin rights, folders, terminal, PATH, proxies |
| ST02 Python | 2 h | Python, virtual environments, pip, Jupyter, VS Code |
| ST03 Colab and Kaggle | 45 min | Notebooks with nothing to install |
| ST04 PostgreSQL | 1.5 h | Server, psql, pgAdmin; load the course data |
| ST05 MySQL | 1.5 h | Server, MySQL Shell, Workbench; load the course data |
| ST06 Git and GitHub | 1 h | Version control for projects |
| ST07 Power BI and Power Query | 1 h | Power BI Desktop and Excel Power Query |
| ST08 Ollama (optional) | 1 h | Local open models |
Questions about a programme?
Tell us your background and goal and we will suggest where to start.