Generative AI
SearchStart here

All paths

ProductionLearning path

Production

Cost, serving, observability, evaluation, fine-tuning and system design.

Lessons
14
Time
8 hours
Modules
3

Skills you will practise

  • Retrieval-augmented generation
  • Agents
  • Safety and guardrails
  • Cost and latency
  • Inference and serving
  • Observability
  • LLM evaluation
  • Production operations
  • Model adaptation choices
  • Fine-tuning
  • GenAI system design
Start the pathCompare paths

You should know

These lessons are not in this path, but some lessons here build on them.

Modules in this path

Open a module to see its lessons.

  1. Production: cost, serving and observability5 lessons 2.5 hours
    1. Cost and latency optimizationProduction 30 min
    2. LLM inference and servingProduction 35 min
    3. Observability and eval pipelinesProduction 30 min
    4. Production deployment of RAGProduction 30 min
    5. RAG security, reliability and releaseProduction 25 min
    About this module
  2. Evaluation and model adaptation4 lessons 2 hours
    1. Evaluating LLM systemsProduction 35 min
    2. Fine-tuning vs RAG vs promptingProduction 25 min
    3. Fine-tuning methods: SFT, LoRA, QLoRA and DPOProduction 25 min
    4. Fine-tuning data, risks and shippingProduction 20 min
    About this module
  3. GenAI system design5 lessons 4 hours
    1. Designing a RAG QA systemProduction 40 min
    2. Designing an agent productProduction 40 min
    3. Designing content moderation and guardrail systemsProduction 40 min
    4. RAG interview drills and mockProduction 60 min
    5. Coding, project defense and mock roundsProduction 45 min
    About this module