Post Graduate Certificate Program inArtificial Intelligence & Machine Learning

• Foundational Python: syntax, variables, data types, operators and more
• Functions, object-oriented programming, file handling, and error handling
• AI-assisted coding with GitHub Copilot: prompt design and code generation
• Reviewing, debugging, optimizing, and automating AI-generated code
• Ethical and legal considerations around AI-generated code

• The data science process and its applications across healthcare, search engines, and finance
• Python packages for data science and NumPy: arrays, indexing, slicing, and arithmetic, statistical, and string operations
• Pandas: Series, DataFrames, statistical operations, and date, timedelta, and categorical data handling
• Sorting, iteration, and text data handling in Pandas
• Data visualization with Matplotlib, Seaborn, and Plotly, including 2D and 3D techniques

• Linear algebra: vectors, matrices, operations, eigenvalues, eigenvectors, and calculus
• Statistics fundamentals and types of data
• Measures of central tendency, dispersion, and shape
• Hypothesis testing, confidence intervals, p-value, and tests including T-Test and more
• Feature engineering: scaling, encoding, and transformations

• ML intro & types; supervised learning basics
• Regression (linear/poly) – evaluation & tuning
• Classification (Logistic, Naive Bayes, KNN, Trees, SVM, RF)
• Imbalanced data; ensemble: bagging, boosting, stacking
• Unsupervised: clustering, PCA/LDA/t-SNE, anomaly detection & recommendations

• Deep Learning vs ML: Neural networks, activation functions, forward/backpropagation & gradient descent
• Deep networks with TensorFlow, Keras & PyTorch; regularization & model building
• Optimization (SGD, Adam, RMSProp), batch norm, dropout & early stopping
• CNNs, transfer learning & object detection (YOLO v3, TF Lite)
• RNNs (LSTM, GRU) & autoencoders for unsupervised learning

• NLP basics: tokenization, stemming, lemmatization & classification (NLTK)
• Text vectorization: BoW, TF-IDF; embeddings: Word2Vec, GloVe
• Machine translation & evaluation metrics
• Seq models: RNNs, Seq2Seq, Transformers, BERT & GPT
• Audio analytics: DSP, Fourier, MFCCs, speech recognition & GAN-based generation

• VAEs, GANs, Transformers & RAG model types
• Transformers use self-attention & multi-head attention
• RAG + LLM architecture & training
• LangChain: models, prompts, memory & chains
• Prompt engineering: zero-shot, few-shot & CoT

• LangChain: model I/O, loaders, splitters, embeddings & vector stores
• Chains, memory & agents with tool integration for conversational/retrieval systems
• LLM fine-tuning: Supervised, PEFT & RLHF with hyperparameter tuning & evaluation
• Fine-tuning via HuggingFace, Accelerate, DeepSpeed & bias mitigation
• Benchmarking with ROUGE, HELM, GLUE, SuperGLUE & BIG-bench

• MLOps lifecycle, benefits, pillars, maturity levels & feature stores
• Version control & experiment tracking with MLflow
• Model deployment architecture, monitoring, metrics & feedback loops
• Automation, orchestration & IaC with Step Functions, CloudFormation & Terraform
• MLOps on cloud, drift detection, security, governance & compliance

• Tool-context pairing and the role of structured context
• MCP vs API and the benefits of protocolized tool access
• Setting up an MCP-compliant tool with FastMCP, including registration, hosting, and discoverability
• System prompt architecture, guardrails, and context persistence strategies
• MCP security risks: authentication, token management, context poisoning, and data leakage

• How RAG solves the knowledge problem by linking agents to real-time information
• Foundry IQ as a shared knowledge platform for multiple agents
• Configuring data sources: Azure AI Search, Blob Storage, SharePoint, and OneLake
• Configuring retrieval behavior and agent instructions for consistent, cited responses
• Integrating, testing, and monitoring agent retrieval in production

The program culminates in a capstone project guided by industry experts, where learners tackle real-world business problems and apply module concepts to build end-to-end solutions strengthening technical skills and creating a portfolio-ready project that showcases job readiness.