Live cohortData & AIANCHOR
Agentic AI & RAG
A 12-week beginner course: build AI assistants that search documents, cite sources, admit when they don't know, and ask for approval before acting.
Cohort starts9 November 2026
Duration12 weeks · ~8 hrs a week
Curriculum6 modules · 28 lessons
LevelBeginner
LanguageEnglish
Fee₹50,000
What you will be able to do
Python and LLM fundamentals: tokens, prompts and structured output
RAG measured with recall@k and faithfulness, using hybrid search and reranking
One deployed agentic assistant with an approval-gated action tool and a 30+ question eval set
A cost report and design note to walk an interviewer through
Who it is for
- People who have never coded or have only tried a little
- A laptop with Chrome: labs run in Google Colab
- Can commit about 8 hours a week for 12 weeks
- Want working assistants, not prompt tricks or a certificate
Curriculum
6 modules, 96 hours
- ANC-01Python for AI BuildersVariables, strings, lists and dictionaries · Functions and JSON · Building a keyword FAQ bot · Your first LLM API call16 h
- ANC-02Talking to LLMsTokens and multilingual cost differences · System prompts, temperature and few-shot examples · Schema-valid JSON output · Per-field accuracy evaluation16 h
- ANC-03Embeddings & Your First RAGEmbeddings for comparing meaning · Chunking documents · Chroma vector database · Prompts that force citations and honest refusals · English and Hindi questions on one index16 h
- ANC-04Better RAG, MeasuredEvaluation sets and recall@k · BM25 keyword search · Hybrid search with rank fusion · Cross-encoder reranking · LLM judge for faithfulness and its biases16 h
- ANC-05Agents: Tools, Loops & MCPHand-written agent loop with step limits and traces · Approval gates for actions · Prompt-injection attack and defence · The same agent in LangGraph · Tools published over MCP16 h
- ANC-06Agentic RAG, Deployment & CapstoneSelf-grading retrieval with query rewriting · Human-in-the-loop pauses · Token and cost logging · FastAPI endpoint · Capstone project review16 h