Data Science with AIML
Take a real, messy problem end to end: frame it, engineer the data, build and rigorously evaluate models, then ship a deployed, monitored data product. This is the on-ramp of the AI family.
Four rungs. Start where you are, finish at Mastery.
Sessions are always 2 hours. The hours are fixed; the calendar flexes — the same level runs daily, alternate-day, or as a summer intensive. Prices shown are student prices; professional pricing is on the pricing page.
What data science actually is
See a real dataset go from mess to answer in one sitting — and find out whether this way of thinking suits you. No coding background needed.
Python, wrangling and EDA
Enough Python to be dangerous, plus Pandas and the exploratory data analysis habits that separate a real analyst from someone running charts.
Classical ML and a real project
Regression, classification, evaluation done honestly, and a full project on real data that you scope, build and defend.
End-to-end, deployed and monitored
Mathematics, ensembles, interpretability, deep-learning basics, MLOps and applied Gen AI — ending in a deployed data product. Full syllabus below.
What the levels below Mastery cover
Our ladder is a spiral, not a straight line. L0 and L1 are outcome-first: you make something impressive on the first day and no theory gets in the way. L2 is where the foundations genuinely begin. L3 then deliberately re-covers 60–70% of L2 at real depth before going past it — the overlap is a feature, not repetition. You earn the "why" only once you are invested in the "what".
A demo and a guided first build. No theory at all — just a win you can point at.
Bigger builds, same spirit. Concepts appear only where a build actually needs them.
The first real grounding in how and why — taught applied, ending in a project.
Re-covers L2's foundations at full depth, then goes well beyond into advanced work.
"I found an insight and ran a prediction — today."
See it: AI predicting prices, classifying images and spotting fraud; a live dashboard; a chatbot answering questions from a spreadsheet.
Make it, guided: Open a ready-made notebook, feed it data and watch it predict something; make one chart that reveals a real insight; ask an AI to analyze a small dataset for you.
Runs in a browser notebook (Colab) — nothing to install. No assessment.
Six sessions, six things you built
- Data to insight — load a real dataset, explore it, make your first chart
- Ask questions of data — filter, group, and a simple dashboard view
- Predict something — a guided ML model predicts a value such as a price or score
- Classify something — spam, sentiment, or a simple image classifier (pretrained)
- AI on your data — use an LLM to analyze and summarize a dataset
- Your own mini analysis on a dataset you choose, plus a showcase
You leave with a small analysis and a working prediction, both shareable. Assessed on completion and the showcase.
L2 · Practitioner — 40 hours, where the foundations begin
The first real grounding — Python for data, how ML actually works, and an end-to-end mini project.
| Unit | What you learn and build | Sessions | Hours |
|---|---|---|---|
| U1 · Python for Data | Python essentials, notebooks, NumPy and pandas basics | 3 | 6 |
| U2 · Data Wrangling & EDA | Cleaning, transforming and exploring; visualization that tells a story | 4 | 8 |
| U3 · Foundations of ML | What ML is, the train/test workflow, regression, evaluation basics | 3 | 6 |
| U4 · Classification & Evaluation | Classifiers (logistic, trees, kNN), metrics, over- and underfitting, cross-validation | 4 | 8 |
| U5 · Applied Taste of More | Clustering intro, a taste of deep learning and NLP via libraries, using AI tools to work faster | 3 | 6 |
| U6 · Practitioner Project | An end-to-end mini ML project — data to model to insight — presented | 3 | 6 |
| Total | 20 | 40 |
Project: An end-to-end mini ML project: raw data through to an evaluated model with documented insights.
Tools: Python · pandas · Matplotlib/Seaborn · scikit-learn · light PyTorch in U5 · Colab/Jupyter · Git.
Assessment: Continuous labs 40% · unit checkpoints 20% · project 40%.
Overlap into L3: U1–U4 are the foundations L3 deepens — it adds maths and statistics rigour, ensembles with tuning and SHAP, MLOps and deployment, deep learning and NLP, applied Gen AI, forecasting, recommenders and a deployed capstone.
120 hours, unit by unit
This is the complete Mastery syllabus — the ceiling of the stream, and what every level below builds toward.
The outcome
You can take a real, messy problem end to end — frame it, engineer the data, build and rigorously evaluate models, interpret them, and ship a deployed, monitored data product — using AI tools to accelerate the workflow rather than replace your judgement.
How it runs
120 hours · 60 sessions × 2 hours · about 3 months. Prerequisite: Data Science L2, or a placement check. No paid cloud account needed — labs run on free tiers and Colab.
Tools & environment
Deliberate stack choices: PyTorch for all deep-learning work, not TensorFlow. AutoML (AutoGluon) is excluded on purpose — you tune models by hand so you understand what AutoML would have automated.
| Unit | What you learn and build | Sessions | Hours |
|---|---|---|---|
| U1 · Math & Statistics | Linear algebra (vectors, matrices, eigen, SVD, PCA foundations) · calculus and gradients · probability and distributions · hypothesis testing, confidence intervals, p-values, A/B testing · CLT, correlation vs causation · entropy and cross-entropy | 5 | 10 |
| U2 · Advanced Python & Data Engineering | Idiomatic Python — comprehensions, generators, decorators, OOP for pipelines · NumPy vectorization · advanced Pandas (groupby, merge, pivot, time-series) · scikit-learn Pipeline and ColumnTransformer · feature engineering · APIs, configs, environments · Git and reproducible experiments | 5 | 10 |
| U3 · Analytics & Visualization | Rigorous EDA · SQL for analytics — joins, windows, aggregation · storytelling visuals · dashboards (Power BI or Streamlit) · two to three business case studies | 4 | 8 |
| U4 · Supervised Learning | Linear and polynomial regression, regularization (Ridge/Lasso/ElasticNet), diagnostics · logistic regression, kNN, decision trees, SVM with kernels, Naive Bayes · metrics — accuracy, precision/recall, F1, ROC-AUC, confusion matrix · imbalanced data (SMOTE, class weights, threshold tuning) · cross-validation and bias-variance | 7 | 14 |
| U5 · Ensembles & Optimization | Bagging and Random Forest · boosting (AdaBoost, Gradient Boosting) · XGBoost, LightGBM, CatBoost · hyperparameter tuning (GridSearch, RandomizedSearch, Optuna) · interpretability with SHAP, LIME and feature importance | 6 | 12 |
| U6 · Unsupervised Learning | K-Means, hierarchical, DBSCAN, GMM · cluster evaluation (silhouette, elbow, Davies-Bouldin) · dimensionality reduction (PCA, t-SNE, UMAP) · anomaly detection (Isolation Forest, One-Class SVM, LOF) · customer segmentation | 4 | 8 |
| U7 · Deep Learning & NLP intro | Neural nets and training in PyTorch · CNN intuition for vision · RNN/LSTM for sequences and time series · the NLP pipeline — tokenization, TF-IDF, embeddings · sentiment and NER · an attention preview that bridges to the Deep Learning stream | 5 | 10 |
| U8 · MLOps & Deployment | Versioning and experiment tracking with MLflow · packaging a model · serving with FastAPI · Docker · cloud deploy · CI/CD with GitHub Actions · monitoring, drift and logging | 6 | 12 |
| U9 · Applied Gen AI for Analysts | Using LLMs in the data workflow · prompt engineering · embeddings and a simple RAG over your own data · responsible AI and hallucination checks · when to build versus call an API | 3 | 6 |
| U10 · Time Series & Forecasting | Time-series data and stationarity · ARIMA and SARIMA · Prophet · feature-based forecasting · backtesting and evaluation | 5 | 10 |
| U11 · Recommenders & Specialized ML | Collaborative and content-based filtering · matrix factorization · evaluating recommenders · imbalanced and anomaly use-cases · choosing the right approach | 5 | 10 |
| U12 · Capstone | Scope → data → model → rigorous evaluation → deployment → stakeholder presentation | 5 | 10 |
| Total | 60 | 120 |
The capstone
A deployed, portfolio-grade data product: a real problem, real data, a validated model, a usable interface or API, and a stakeholder presentation.
Three tracks to choose from — an applied ML system, an analytics and dashboard product, or a RAG-assisted data assistant.
You deliver: a live deployment, a repository, an evaluation write-up and a presentation to a non-technical stakeholder.
How you're assessed
Syllabus is a working draft — hours per unit are indicative and validated by the practitioner who teaches the stream. Reinforcement learning is deliberately out of scope for the analyst path.
Who will teach you
Every instructor is currently building in the field they teach — and is certified on our level template before taking a cohort alone. Subject expertise is the entry ticket, not the job.
Practitioners, not lecturers
Every instructor is currently building in the field they teach. If they stop practising, they stop teaching that stream.
Trained on our template
Subject knowledge is the entry ticket, not the job. Every instructor is certified on the CEFTA level template before they take a cohort alone.
Standardized delivery
The same module runs the same way in every centre and every partner campus. That consistency is the product.
Named instructor profiles go up as each cohort's trainer is confirmed — we publish people, not stock photographs. We're hiring practitioner-instructors →
Tell me when the next batch opens
One email when a cohort is published for this stream. Nothing else, ever.
Turn messy data into something that ships.
Start with the free 3-hour L0 Discover session — see the whole workflow once, then decide whether to climb.