Five pipeline and batch patterns. Which one was the most useful insight?
That AI agents are just functions — batch processing is just a list comprehension. I kept expecting a special 'batch mode'. It's [classify_urgency(t) for t in texts].
That insight unlocks the rest of the track. The AI part is always a single function call. The scale comes from Python iteration — nothing special about the AI layer. What was confusing?
min(key=len) — I kept mixing up when to use key=len vs key=lambda x: len(x). And whether the list comprehension inside batch_word_counts can call .split() inline.
key=len and key=lambda x: len(x) are equivalent — len is a function object, so you can pass it directly as key. Six questions — four from this week, two from Weeks 1 and 2.
Ready. Week 4's web search is what I've been building toward — a research assistant that actually searches the web.
The patterns you built this week are the foundation for that capstone.
Five pipeline and batch patterns. Which one was the most useful insight?
That AI agents are just functions — batch processing is just a list comprehension. I kept expecting a special 'batch mode'. It's [classify_urgency(t) for t in texts].
That insight unlocks the rest of the track. The AI part is always a single function call. The scale comes from Python iteration — nothing special about the AI layer. What was confusing?
min(key=len) — I kept mixing up when to use key=len vs key=lambda x: len(x). And whether the list comprehension inside batch_word_counts can call .split() inline.
key=len and key=lambda x: len(x) are equivalent — len is a function object, so you can pass it directly as key. Six questions — four from this week, two from Weeks 1 and 2.
Ready. Week 4's web search is what I've been building toward — a research assistant that actually searches the web.
The patterns you built this week are the foundation for that capstone.
Create a free account to get started. Paid plans unlock all tracks.