Track 04 · 15 lessons
AI Product Management.
Choose where AI helps, and design for where it fails.
What an LLM Actually Does
Understand why useful generation is not the same as verified knowledge.
Scroll lessonTokens, Context and Sampling
Connect model settings to memory, variation, cost, and user expectations.
Scroll lessonChoosing a Model
Select for the task and its failure cost, not the leaderboard.
Scroll lessonPrompt and Instruction Design
Make the intended behavior explicit and test changes like product changes.
Scroll lessonStructured Outputs
Separate predictable format from correct meaning.
Scroll lessonTool Calling
Understand how a model's proposed action becomes a software operation.
Scroll lessonRAG
Use retrieval to supply evidence, then evaluate both finding and answering.
Scroll lessonEmbeddings and Vector Search
Use semantic similarity without confusing it with exactness or truth.
Scroll lessonAgents and Workflows
Choose autonomy only where adapting the next step creates value.
Scroll lessonEvals
Define quality before the demo persuades you the system is ready.
Scroll lessonHallucinations and Uncertainty
Design around the consequences of plausible but unsupported output.
Scroll lessonHuman in the Loop
Put review where a person can actually detect and prevent the important error.
Scroll lessonAI Cost, Latency and UX
Optimize the time and money spent on a successful user outcome.
Scroll lessonAI Safety, Privacy and Data
Ask where information goes and what the system is allowed to do with it.
Scroll lessonWhen NOT to Use AI
Choose the simplest method that can meet the task's correctness requirements.
Scroll lesson