🤖 AI & Agentic AIBeginner LevelEvidence-Based Skill Profile
Hands-on Mastery in RAG & AI Knowledge Systems
Don't just store knowledge. Build AI systems that can retrieve, understand, and use the right knowledge. Master high-throughput document ingestion, intelligent semantic chunking, dense and sparse vector representations, vector database indexing, hybrid retrieval with Reciprocal Rank Fusion, cross-encoder reranking, context window budgeting, citation-grounded generation, RAG evaluation frameworks (Ragas / LLM-as-a-judge), and defense-in-depth security guardrails with your dedicated in-course RAG & Knowledge Systems Mastery Agent coach.
14 Modules • 7 Production Projects
14 Modules
AI FearFilter Faculty
What You Will Learn
Architect end-to-end Retrieval-Augmented Generation (RAG) pipelines bridging private knowledge stores with foundation models.
Ingest, parse, and clean unstructured multi-modal documents (PDFs, Markdown, HTML, tables) while preserving semantic hierarchy.
Implement optimal chunking strategies including recursive character splitting, sliding windows, and Parent-Document retrieval.
Generate and evaluate dense, sparse, and multi-vector embeddings using cosine similarity and dot product metrics.
Deploy and configure production vector databases (pgvector, Chroma, Qdrant) using HNSW and IVF approximate nearest neighbor indexes.
Build hybrid retrieval engines combining BM25 keyword search with dense vector similarity via Reciprocal Rank Fusion (RRF).
Formulate intelligent query transformations including HyDE (Hypothetical Document Embeddings) and sub-query routing.
Apply cross-encoder reranking models to filter low-relevance passages and maximize precision@k.
Construct defensive context windows with bracketed source citations and strict refusal mechanics when evidence is insufficient.
Evaluate RAG accuracy and eliminate hallucinations using automated benchmarks measuring Faithfulness and Context Precision.
Enforce strict server-side tenant isolation, document ACL checks, and guardrails against indirect prompt injection in retrieved data.
Build 7 production portfolio projects and pass the Module 14 Capstone Challenge evaluated by the RAG & Knowledge Systems Mastery Agent.
Curriculum & Weekly Roadmap
14 Structured ModulesModule 1 — Introduction to RAG & AI Knowledge Systems
- Parametric vs Non-Parametric Memory: Why RAG is Essential
- The Standard RAG Architecture: Ingestion, Retrieval & Synthesis
- RAG vs Fine-Tuning: Latency, Cost, Update Speed & Hallucination Tradeoffs
Module 2 — Knowledge Sources & Document Processing
- Document Parsing Strategies: Unstructured PDFs, HTML & Tabular Data
- Layout-Aware Extraction: Preserving Headers, Columns & Visual Hierarchies
- Metadata Extraction & Breadcrumb Tagging for Governed Enterprise Search
Module 3 — Document Loading, Cleaning & Chunking
- Chunking Tradeoffs: Small Chunks vs Large Chunks & Overlap Tuning
- Recursive Character Splitting & Semantic Boundary Detection
- Parent-Document & Small-to-Big Retrieval Architectures
Module 4 — Embeddings & Semantic Representation
- Dense Embeddings & High-Dimensional Latent Spaces
- Mathematical Vector Distance Metrics: Cosine Similarity, Dot Product & Euclidean Distance
- Sparse Representations: BM25, SPLADE & Multi-Vector ColBERT
Module 5 — Vector Databases & Similarity Search
- Exact KNN vs Approximate Nearest Neighbor (ANN) Indexing
- HNSW (Hierarchical Navigable Small World) & IVF Index Structures
- Production Vector Databases: Dedicated (Chroma/Qdrant) vs Relational pgvector
Module 6 — Retrieval Strategies & Hybrid Search
- The Power of Hybrid Search: Combining Lexical Keywords with Semantic Vectors
- Reciprocal Rank Fusion (RRF): Merging Unbounded Scores via Ordinal Ranks
- Weighted Linear Combinations vs Rank-Based Merging
Module 7 — RAG Pipelines & Context Construction
- Context Assembly & The "Lost in the Middle" Phenomenon
- Chunk Reordering & Token Budget Optimization in Context Windows
- Defensive Prompt Framing: Treating Context Strictly as Passive Reference Data
Module 8 — Query Transformation, Reranking & Retrieval Quality
- Query Rewriting & HyDE (Hypothetical Document Embeddings)
- Multi-Query Decomposition & Step-Back Prompting
- Two-Stage Retrieval: Bi-Encoder Search followed by Cross-Encoder Reranking
Module 9 — Grounded Generation & Citation-Based Answers
- Sentence-Level Citation Mapping & Attribution Verification [Doc X]
- Grounded Refusal Mechanics: Saying "I Don't Know" When Facts are Missing
- Post-Generation Natural Language Inference (NLI) Verification
Module 10 — RAG Evaluation, Accuracy & Hallucination Reduction
- The RAG Triad: Context Relevance, Groundedness (Faithfulness) & Answer Relevance
- Automated CI/CD Benchmarking with Ragas & LLM-as-a-Judge
- Detecting and Mitigating Retrieval Regressions
Module 11 — Advanced RAG & Production Knowledge Systems
- Agentic RAG: Dynamic Query Routing, Tool Invocation & Iterative Multi-Hop Retrieval
- Self-RAG & Corrective RAG (CRAG): Dynamic Retrieval Triggers
- GraphRAG: Combining Knowledge Graphs with Vector Embeddings
Module 12 — RAG Security, Privacy & AI Guardrails
- Retrieved Documents are DATA, Not Instructions: Defending Against Indirect Prompt Injection
- Server-Side Multi-Tenant Isolation & Document-Level Access Control Lists (ACLs)
- DPDP & GDPR Compliance: Real-Time Vector Deletion & Data Redaction
Module 13 — Building Production-Ready RAG Applications (7 Projects)
- Projects 1-3: Document Q&A with Citations, Ingestion Pipeline, Hybrid RRF Engine
- Projects 4-5: Company Knowledge Base Assistant & Multi-Document Research Assistant
- Projects 6-7: Automated RAG Eval Benchmark Harness & Production Guarded RAG Service
Module 14 — Final Hands-on RAG & AI Knowledge System Challenge
- Production Enterprise RAG & Knowledge Architecture Specifications
- Implementation: Hybrid Retriever, Citation Generator, Tenant Filter & Guardrails
- Submission, Multi-Criteria Evaluation & RAG Mastery Agent Verification
Who This Course Is For
Software Engineers, AI Engineers, Backend Developers, Data Engineers, Solutions Architects, and Computer Science Students seeking to engineer enterprise-grade retrieval-augmented generation systems, enterprise search engines, and production AI knowledge architectures.
Key Skills Developed:
RAG Architecture & FoundationsDocument Ingestion & Multi-Format ParsingDocument Cleaning & Semantic ChunkingEmbeddings & Semantic Vector SpacesVector Databases & Indexing Strategies (HNSW / IVF)Hybrid Search & Reciprocal Rank Fusion (RRF)RAG Pipelines & Context ConstructionQuery Transformation & Multi-Query RoutingCross-Encoder Reranking & Retrieval QualityGrounded Generation & Citation AttributionRAG Evaluation & Hallucination Metrics (Ragas)Advanced Agentic & Graph-Based RAGRAG Security, Tenant Isolation & GuardrailsProduction Capstone Enterprise RAG Challenge
Course Faculty & Development
AI FearFilter Faculty
Retrieval Systems & Knowledge Engineering Team
AI FearFilter Academy
Engineering CurriculumAI FearFilter Academy
100% FREEFree For All Students
100% Self-Paced + Active Hands-on Learning
Evidence-Based Demonstrated Skill Profile
Full Lifetime Access in Student Home
FILTER FEAR. TRUST FACTS.