Designing Generative AI Systems: Architecture, Scale, and Deployment

Advanced22 hours10 sections42 lessonsRated 4.9 out of 5 from 28 ratings

Design generative AI services around explicit requirements, authorized retrieval, capacity estimates, evaluation, safety checks, and reversible releases. Text, image, speech, video, and captioning cases show how the tradeoffs change by workload.

What you will be able to do

  • Choose between retrieval and fine-tuning, and estimate training and serving resources.
  • Design training and serving paths for text, image, speech, and video generation.
  • Apply evaluation, safety checks, and inference optimizations to generative workloads.
  • Compare agentic architectures with fixed generation services.

Who this course is for

Engineers preparing for generative AI system design interviews or building production generative AI services.

Helpful to know first

  • How large language models work, and system design basics such as caching, queues, and capacity estimates.

Course content

10 sections 42 lessons 22 hours

Choose evaluation metrics and compare parallelism and inference optimization techniques. Review the core concepts used throughout the system design cases.

  1. Evaluation Metrics for GenAI Systems15 min
  2. Parallelism in GenAI Models13 min
  3. Inference Optimization in GenAI Systems10 min
  4. Key Concepts in Designing GenAI Systems13 min

Locked sections open with Premium.Upgrade to Premium

Rate this course

Rated 4.9 out of 5 from 28 ratings

Tell us how useful this course was. You can update your rating later.

Loading your rating...

Course rating

Rated 4.9 out of 5 from 28 ratings

  • 5 stars: 25 ratings
  • 4 stars: 3 ratings
  • 3 stars: 0 ratings
  • 2 stars: 0 ratings
  • 1 star: 0 ratings