This is the AI LaunchPad track syllabus — 6 months · 24 weeks.View the full program
What you'll be able to do
- Clean and analyze real datasets with Pandas and NumPy
- Visualize data and run EDA with Matplotlib and Seaborn
- Build and evaluate machine-learning models with scikit-learn
- Apply NLP basics and understand the deep-learning workflow
- Deploy a trained model and tune it for better performance
Tools & technologies
PythonpandasNumPyscikit-learnJupyterPandasMatplotlibRegressionClassificationClusteringModel Evaluation
The syllabus
Python and Programming Foundations
Python Fundamentals and Initial Logic Building
- Day 1Computer and programming fundamentals · Python installation · VS Code and Jupyter setup · terminal basics · first Python programBuild: Interactive Student Profile
- Day 2Variables · identifiers · Python data types · input and output · type conversionBuild: Smart Bill Calculator
- Logic Building 1Problem decomposition · sequence-based problems · pseudocode · flowcharts · arithmetic logicBuild: Multi-Unit Converter
- Day 4Operators · strings · indexing · slicing · string methods · formatted stringsBuild: Smart Text Formatter
- Day 5Boolean expressions · comparison operators · if, elif, else · nested conditionsBuild: Eligibility Decision Engine
- Logic Building 2Decision tables · multi-condition problems · boundary cases · branch debuggingBuild: Dynamic Ticket-Pricing System
Loops, Collections and Advanced Logic Building
- Day 7while loops · counters · accumulators · sentinel-controlled loops · loop validationBuild: ATM PIN and Withdrawal Simulator
- Day 8for loops · range · nested loops · loop control · break and continueBuild: Number Pattern Generator
- Logic Building 3Loop-based problem solving · digit manipulation · divisibility · prime numbers · pattern logicBuild: Number Challenge Toolkit
- Day 10Lists · tuples · indexing · slicing · list methods · iteration · list comprehensionsBuild: Student Marks Manager
- Day 11Dictionaries · sets · key-value operations · membership · frequency countingBuild: Inventory Lookup System
- Logic Building 4Collection-based problems · searching · counting · duplicate detection · menu-driven programsBuild: Console Contact Book
Functions, Files, OOP and Numerical Python
- Day 13Functions · parameters · arguments · return values · scope · reusable codeBuild: Python Utility Library
- Day 14Text and CSV file handling · exceptions · debugging · validation · error messagesBuild: File-Based Expense Tracker
- Day 15Classes · objects · constructors · methods · inheritance basics · modules and packagesBuild: Library Management System
- Day 16Command line · Git fundamentals · GitHub repositories · commits · branches · README · project structureBuild: Version-Controlled Python Portfolio
- Day 17NumPy arrays · dimensions · shapes · data types · indexing · slicing · reshapingBuild: Numerical Array Toolkit
- Day 18Vectorisation · broadcasting · aggregate operations · matrix operations · random arraysBuild: Image Pixel Transformer
ML Data and Mathematics Foundations
Data Preparation Required for Machine Learning
- Day 19Pandas Series · DataFrames · CSV loading · dataset shape · columns · data typesBuild: Dataset Loader and Inspector
- Day 20Row and column selection · filtering · sorting · aggregation · groupingBuild: Student Performance Filter
- Day 21Missing values · duplicate records · incorrect types · outlier detection · data-quality checksBuild: Automated Dataset Cleaner
- Day 22Numerical and categorical features · encoding · scaling · feature-target separationBuild: ML Feature Transformer
- Day 23Matplotlib fundamentals · feature distributions · relationships · class balance · model-error plotsBuild: ML Visual Diagnostics Tool
- Day 24Reusable preprocessing · train-test consistency · transformations · preprocessing pipelinesBuild: ML Ready Data Pipeline
Intuitive Mathematics for Machine Learning
- Day 25Scalars · vectors · matrices · tensors · shapes · transpose · element-wise operationsBuild: Vector and Matrix Calculator
- Day 26Dot products · matrix multiplication · norms · Euclidean distance · cosine similarityBuild: Similarity Search Engine
- Day 27Probability fundamentals · events · conditional probability · independence · Bayesʼ theoremBuild: Bayesian Risk Calculator
- Day 28Mean · median · variance · standard deviation · distributions · covariance · correlationBuild: Statistical Feature Profiler
- Day 29Functions · slope · derivatives · partial derivatives · gradients · chain ruleBuild: Numerical Gradient Checker
- Day 30Loss functions · optimisation · gradient descent · learning rate · convergenceBuild: From-Scratch Gradient Descent Optimiser
Core Machine Learning and Supervised Learning
Machine Learning Workflow
- Day 31Artificial intelligence · machine learning · deep learning · rule-based systems · supervised, unsupervised and reinforcement learningBuild: AI Use-Case Mapper
- Day 32ML problem framing · inputs · features · target · constraints · success criteria · business-to-ML translationBuild: ML Problem Statement Canvas
- Day 33Training, validation and test sets · random splitting · stratification · baselines · data leakageBuild: Leakage-Safe Dataset Splitter
- Day 34Scikit-learn estimator API · fit · predict · transform · model parameters · model attributesBuild: Baseline Prediction System
- Day 35Pipelines · ColumnTransformer · preprocessing consistency · reproducibility · random statesBuild: Reusable Scikit-Learn Pipeline
- Day 36Complete ML workflow · training · evaluation · model persistence · loading · inferenceBuild: First End-to-End ML Predictor
Regression
- Day 37Regression problems · simple linear regression · line of best fit · coefficients · predictionBuild: House Price Line-Fit Visualiser
- Day 38Linear regression cost function · gradients · parameter updates · training from scratchBuild: From-Scratch Salary Predictor
- Day 39Multiple linear regression · multiple features · coefficients · multicollinearity conceptsBuild: Used-Car Price Predictor
- Day 40Feature scaling · polynomial features · interaction features · nonlinear relationshipsBuild: Nonlinear Rent Predictor
- Day 41Overfitting · underfitting · Ridge · Lasso · Elastic Net · regularisation strengthBuild: Regularised Property Valuation
- Day 42MAE · MSE · RMSE · R² · residuals · cross-validation · regression comparisonBuild: Regression Benchmark System
Classification
- Day 43Classification problems · logistic regression · sigmoid function · probabilities · log lossBuild: Admission Probability Classifier
- Day 44Decision boundaries · thresholds · binary and multiclass classification · one-vs-restBuild: Classification Threshold Tuner
- Day 45K-nearest neighbours · distance metrics · feature scaling · selecting KBuild: Species Recognition Classifier
- Day 46Naive Bayes · conditional probability · Gaussian, Multinomial and Bernoulli variantsBuild: Spam Message Detector
- Day 47Support Vector Machines · margins · support vectors · kernels · C and gammaBuild: Tumour Classification System
- Day 48Confusion matrix · precision · recall · F1 · ROC AUC · PR AUC · class imbalance · calibrationBuild: Rare-Fraud Classification System
Trees, Ensembles and Model Improvement
- Day 49Decision trees · nodes · branches · impurity · entropy · Gini index · information gainBuild: Loan Approval Decision Tree
- Day 50Tree depth · minimum samples · overfitting · pruning · feature importanceBuild: Pruned Credit-Risk Model
- Day 51Bagging · bootstrap sampling · random forests · random feature selection · out-of-bag evaluationBuild: Credit Default Random Forest
- Day 52Gradient boosting · sequential learning · learning rate · XGBoost fundamentalsBuild: Employee Attrition Booster
- Day 53Grid search · randomised search · search spaces · cross-validated tuning · model selectionBuild: Automated Model Tuner
- Day 54Bias and variance · learning curves · error analysis · feature importance · SHAP · fairness · model cardsBuild: Explainable Credit-Scoring System
Unsupervised and Specialised Machine Learning
Unsupervised Learning, Anomaly Detection and Recommendations
- Day 55Clustering problems · K-means · centroids · assignments · initialisation · selecting KBuild: Customer Grouping System
- Day 56Hierarchical clustering · dendrograms · linkage methods · DBSCAN · density-based clustersBuild: Location Hotspot Detector
- Day 57Silhouette score · inertia · cluster stability · Gaussian mixture models · cluster interpretationBuild: Cluster Quality Evaluator
- Day 58Dimensionality reduction · PCA · explained variance · reconstruction · SVDBuild: Digit Feature Compressor
- Day 59Anomaly detection · statistical thresholds · Isolation Forest · One-Class SVMBuild: Machine Failure Anomaly Detector
- Day 60User-item matrices · collaborative filtering · content-based filtering · similarity · ranking · evaluation at KBuild: Movie Recommendation Engine
Neural Networks and Applied Artificial Intelligence
Neural Network Foundations with PyTorch
- Day 61Biological neuron analogy · perceptron · weights · bias · activation functions · network layersBuild: Logic-Gate Neural Unit
- Day 62PyTorch tensors · tensor shapes · devices · datasets · DataLoaders · GPU conceptsBuild: PyTorch Tensor Operations Lab
- Day 63Input, hidden and output layers · forward propagation · logits · probabilities · loss selectionBuild: Fashion-Item Neural Network
- Day 64Computational graphs · backpropagation · gradients · autograd · gradient checkingBuild: Backpropagation Checker
- Day 65SGD · momentum · Adam · mini-batches · epochs · training loops · validation loops · checkpointsBuild: Reusable Neural Training Engine
- Day 66Tabular neural networks · training · validation · metrics · prediction · saving and loadingBuild: Neural Classification System
Deep Learning and Computer Vision
- Day 67Neural-network overfitting · weight decay · dropout · batch normalisation · early stopping · learning-rate schedulingBuild: Regularised Neural Classifier
- Day 68Images as tensors · channels · kernels · convolution · feature maps · padding · stride · poolingBuild: Convolution Visualiser
- Day 69CNN architecture · convolution blocks · flattening · dense layers · image-classification outputsBuild: CNN Digit Recogniser
- Day 70Image loading · resizing · normalisation · augmentation · train-validation image pipelinesBuild: Augmented Object Classifier
- Day 71Pretrained CNNs · transfer learning · frozen layers · feature extraction · fine-tuningBuild: Plant-Disease Image Classifier
- Day 72CNN evaluation · confusion matrix · per-class metrics · inference pipeline · checkpoint managementBuild: End-to-End Image Classification System
Traditional NLP and Sequence Models
- Day 73Text cleaning · sentence splitting · tokenisation · stop words · stemming · lemmatisationBuild: Text Preprocessing Engine
- Day 74Bag of words · n-grams · TF IDF · sparse matrices · text similarityBuild: Document Similarity Finder
- Day 75Logistic regression · Naive Bayes · SVM for text · text-classification pipelinesBuild: News Category Classifier
- Day 76Word embeddings · vector representations · similarity · embedding layers · vocabulary handlingBuild: Word Similarity Explorer
- Day 77Sequence data · recurrent neural networks · hidden states · LSTM · GRUBuild: Sequence Sentiment Model
- Day 78Sequence padding · masking · batching · text-model evaluation · neural text inferenceBuild: Product Review Sentiment System
Time-Series Machine Learning and Reinforcement Learning
- Day 79Temporal ordering · trends · seasonality · lag features · rolling features · date-time featuresBuild: Demand Forecast Feature Builder
- Day 80Forecasting baselines · temporal splits · walk-forward validation · MAE · RMSE · MAPEBuild: Walk-Forward Forecast Validator
- Day 81Regression models for forecasting · tree-based forecasting · recursive and direct predictionBuild: Traffic Volume Forecaster
- Day 82Sequence windows · LSTM forecasting · multi-step prediction · sequence-model evaluationBuild: Energy Demand LSTM Forecaster
- Day 83Reinforcement learning · agents · environments · states · actions · rewards · policies · Markov decision processesBuild: Grid-World Environment
- Day 84Q-learning · Q-tables · exploration and exploitation · epsilon-greedy policy · agent evaluationBuild: Q Learning Navigation Agent
ML Deployment and Production Foundations
Model Deployment
- Day 85Model serialisation · preprocessing artefacts · versioned model files · batch inference · online inferenceBuild: Versioned Inference Package
- Day 86FastAPI fundamentals · routes · request methods · Pydantic schemas · prediction endpointsBuild: Machine-Learning Prediction API
- Day 87Streamlit fundamentals · user inputs · model loading · prediction display · local application structureBuild: Interactive ML Web Application
- Day 88Containers · Dockerfiles · dependency management · image building · ports · container executionBuild: Dockerised ML Service
- Day 89Input validation · exception handling · structured logging · unit tests · API tests · model testsBuild: Tested and Logged Prediction API
- Day 90Cloud deployment · environment variables · secrets · public endpoints · health checks · deployment verificationBuild: Public Cloud ML Endpoint
MLOps Foundations and Sprint 1 Technical Capstone
- Day 91Experiments · runs · parameters · metrics · artefacts · MLflow trackingBuild: Tracked Experiment Suite
- Day 92Model registry · model versions · staging · promotion · rollback · metadataBuild: Versioned Model Registry
- Day 93Latency · throughput · error rate · input drift · prediction drift · performance monitoringBuild: ML Drift-Monitoring Dashboard
- Day 94Automated tests · GitHub Actions · container builds · deployment workflows · release tagsBuild: Automated ML Delivery Pipeline
- Day 95Kubernetes pods · deployments · services · replicas · resource limits · rolling updatesBuild: Kubernetes ML Prediction Service
- Day 96Training pipeline · model registry · API · Docker · cloud deployment · monitoring · documentation · technical demonstrationBuild: Sprint 1 Production Mini-Capstone