Yashotantra Academy logo

LLM Fundamentals & Vector RAG Systems

Master the modern AI stack: self-attention mechanics, RoPE embeddings, ChromaDB chunking, hybrid BM25 + dense search, cross-encoder reranking, and evaluation harnesses.

0 of 4 chapters passed

75%+ required per chapter quiz to verify competence.

Start Track

1 · Transformers & Attention

Self-attention matrices, query/key/value projections, multi-head attention, and RoPE.

  1. 1. Inside Self-Attention & RoPE MechanicsThe query-key-value dot product, causal masks, and Rotary Position Embeddings.
    10m · 2 MCQs

2 · Embedding Spaces & Chunking

Semantic vectors, cosine similarity, recursive character chunking, and metadata filtering.

  1. 2. Embedding Geometries & Chunking StrategiesDense vectors, cosine distance, semantic chunking, and sliding window overlaps.
    9m · 2 MCQs

3 · Advanced RAG Architectures

Hybrid search (BM25 + Dense), cross-encoder reranking, and contextual compression.

  1. 3. Hybrid Search & Cross-Encoder RerankingCombining sparse BM25 with dense vectors, Reciprocal Rank Fusion (RRF), and Cohere rerankers.
    11m · 2 MCQs

4 · RAG Evaluation & Ragas

Faithfulness, answer relevance, context recall, and automated eval benchmarks.

  1. 4. RAG Triad Evaluation & Benchmark HarnessesFaithfulness, Answer Relevance, Context Precision, and the Ragas evaluation framework.
    10m · 2 MCQs

Explore Complementary Tracks

Artificial IntelligenceBlockchain & Web3Cloud ComputingData Science & EngineeringSystem Design & Distributed SystemsDevOps & Cloud OrchestrationMLOps & Model Lifecycle EngineeringPrompt Engineering & Deterministic AgentsSmart Contract Auditing & SecurityZero-Knowledge Proofs & CryptographyForward Deployed & Full-Stack AISalesforce Admin & Developer Career