Model Online / On-Site
Duration
Modules 6 modules

Course Objective & Scope

In legal, financial, HR and customer operations, an enterprise assistant that fabricates a policy detail or cites a passage that does not support its claim destroys user trust. This course focuses on the generation and verification side of RAG: suppressing parametric priors in favour of retrieved evidence, producing sentence-level verifiable citations, handling conflicting documents and abstaining cleanly when evidence is absent.

Participants build a grounded generation layer using structured citation schemas, post-generation claim verification (NLI and grounding models) and source recency/authority resolution.

Who Should Attend?

  • Software and AI engineers building high-trust, auditable RAG applications
  • Solution architects and data scientists designing knowledge assistants in regulated industries
  • Product and platform teams accountable for response accuracy and source attribution
  • Prerequisite: basic RAG architecture knowledge and familiarity with LLM APIs/Structured Outputs

Key Takeaways

  • Design structured outputs that link every claim to a document ID, page number and verbatim quote span
  • Implement calibrated abstention so the assistant declines to guess when retrieved context is insufficient
  • Handle conflicting or superseded documents using effective-date metadata and explicit conflict reporting
  • Verify answer faithfulness automatically using NLI models, grounding detectors or LLM-as-a-judge
  • Build post-processing checks that eliminate fabricated quotes and invalid source references

Curriculum

Anatomy of Grounding and RAG Hallucination Modes
  • Extrinsic fabrication, intrinsic distortion and citation hallucination
  • Overriding parametric model priors when internal policies contradict public knowledge
Citation Architectures and Verifiable Attribution
  • Inline `[Doc-1, p.4]` markers, structured JSON citation schemas and native provider citation APIs
  • The quote-then-answer pattern: extracting verbatim supporting spans before synthesis
Calibrated Abstention and Conflicting Source Resolution
  • Decision flows for partial evidence, ambiguous queries and unanswerable questions
  • Resolving contradictions using effective dates, version metadata and source authority
Post-Generation Verification and Grounding Checks
  • Decomposing answers into atomic claims and checking entailment with NLI / grounding models
  • Deterministic substring/fuzzy verification that cited quotes actually exist in the source chunk
UX Patterns for Source Verification and User Trust
  • Interactive source cards and passage/bounding-box highlighting in the original PDF
  • Communicating confidence, partial coverage and source conflicts clearly
Hands-On: High-Assurance Cited Answer Pipeline
  • Testing against adversarial, unanswerable and conflicting policy questions
  • Measuring faithfulness, citation precision and citation recall

Other Courses in This Category

  • Advanced RAG Advanced RAG

    Enterprise RAG architectures covering BM25-dense hybrid retrieval, ColBERT, neural reranking, self-correcting Agentic RAG, GraphRAG, Text-to-SQL, and GPU-accelerated serving.

    RAG & Enterprise Knowledge Systems Corporate, Individual
    Live Course
    10 modules
    View
  • Hybrid Search and Reranking for High-Precision RAG

    Boosting RAG retrieval precision and recall by combining BM25/SPLADE lexical search with dense vectors via RRF, query rewriting (multi-query, HyDE) and cross-encoder or ColBERT reranking.

    RAG & Enterprise Knowledge Systems Corporate
    Live Course
    6 modules
    View
  • Metadata Architecture and Permission-Aware Retrieval in RAG

    Designing metadata schemas, dynamic filter extraction and RBAC/ABAC/ReBAC security-trimmed retrieval so enterprise RAG assistants never leak unauthorized documents.

    RAG & Enterprise Knowledge Systems Corporate
    Live Course
    6 modules
    View
  • Agentic RAG: Multi-Step, Corrective and Adaptive Retrieval Architectures

    Building Agentic RAG systems with LangGraph that plan sub-queries, route across heterogeneous data sources, grade retrieved evidence, self-correct queries (CRAG/Self-RAG) and bound latency.

    RAG & Enterprise Knowledge Systems Corporate
    Live Course
    6 modules
    View
  • GraphRAG and Knowledge-Graph-Enhanced Retrieval Systems

    Combining knowledge graphs (Neo4j, property graphs), entity-relationship extraction, community summaries (Microsoft GraphRAG) and vector search to answer multi-hop and corpus-wide enterprise queries.

    RAG & Enterprise Knowledge Systems Corporate
    Live Course
    6 modules
    View
  • RAG Evaluation and Quality Assurance: Golden Datasets, LLM-as-a-Judge and CI/CD

    Replacing ”vibe checks” in RAG with systematic evaluation: golden dataset curation, synthetic test generation, RAGAS/DeepEval/Promptfoo pipelines, CI/CD quality gates and production monitoring.

    RAG & Enterprise Knowledge Systems Corporate
    Live Course
    6 modules
    View

Get in touch