Post Graduate Program in Generative AI and ML

This module builds core Python programming skills, advanced data structures, asynchronous patterns, and numerical/data analysis and visualization using NumPy, Pandas, Matplotlib, and Seaborn, then extends into building AI APIs, backends, and AI-native applications, along with AI pair-programming fundamentals using GitHub Copilot.

This module covers generative AI and LLM foundations, transformer architectures, working with LLM APIs, SDKs, and open-source models, and progresses through prompt engineering essentials and advanced techniques to context engineering, structured outputs, function calling, tool use, and prompt optimization and evaluation with DSPy.

This module covers embeddings, semantic search, and vector databases, developing basic and advanced RAG systems including hybrid retrieval, re-ranking, and agentic RAG, building LLM apps with LangChain and LlamaIndex, multimodal LLMs, deploying end-to-end GenAI applications, securing them with guardrails against prompt injection, evaluating GenAI applications, and fine-tuning with PEFT.

This module covers agentic AI foundations and agent architectures, LangChain core concepts, agents, and tool use, LangGraph for stateful workflows, routing, cycles, human-in-the-loop, and persistence, multi-agent orchestration with CrewAI and Microsoft AutoGen, agentic RAG and GraphRAG for agents, and deep agents with reflection, planning, and long-term memory.

This module covers the Model Context Protocol's architecture and custom servers, MCP ecosystem integrations, agent interoperability via A2A, ACP, and ANP protocols, evaluation and tracing with LangSmith, AI guardrails and safety with NeMo and Guardrails AI, fine-tuning and agent performance optimization, and Dockerizing and deploying AI agents.

This module covers building agentic workflows and automation without code, using n8n, Zapier, Make, and Flowise for no-code agentic AI.

This module covers the foundations of MLOps and LLMOps, LLM infrastructure, tooling, and the open-source stack, deployment, containerization, and scaling of LLM systems, and monitoring, governance, and responsible AI in production.

This capstone module has learners design, create, and deploy an end-to-end agentic AI application, integrating the skills acquired across the programme into a production-grade system.

This self-paced elective covers vibe coding fundamentals and AI-driven development, and AI-powered and AI-native software development using tools such as Cursor AI, Google Antigravity, and Amazon Q Developer.

This self-paced elective covers NLP foundations and text processing, feature engineering and text representation, tokenization and embeddings, sentiment analysis and text classification, neural language models and sequence modelling, and transformers as the path to large language models.

This self-paced elective introduces Claude and the Anthropic model family, the Anthropic API for messages, system prompts, and tool use, Claude Code and agentic development with Claude, and using Claude with MCP.