AI

ESIA

School of Artificial Intelligence

Worldwide cohort
Students across time zones
HomeMaster’sTrainingsProjectsResearchBlogAboutContact
← Back to Trainings

GenAI: LLM Apps (RAG, Agents & Evaluation)

Build reliable LLM applications: RAG pipelines, evaluation, safety, and agent workflows for real delivery constraints.

GenAILLMRAGAgentsEvaluationSafety

Duration

3 days

Format

Remote

Level

Intermediate

Key outcomes

  • Design a production-style RAG pipeline with good chunking/retrieval
  • Reduce hallucinations with grounding and citation patterns
  • Evaluate answer quality and detect failure modes
  • Implement safety basics: prompt injection awareness and guardrails

Syllabus

Day 1 — RAG foundations

  • Chunking strategies and embeddings
  • Retrieval: top-k, reranking (concept), filtering
  • Prompting patterns for grounded answers + citations

Day 2 — Reliability & evaluation

  • Test sets and automated evaluation patterns
  • Hallucination, missing context, bad retrieval
  • Prompt injection: risks and mitigation patterns

Day 3 — Agents & orchestration

  • Tool calling and multi-step workflows
  • Latency/cost trade-offs and caching
  • Monitoring, feedback loops, and iteration strategy