CourseAI System Design: The Complete Course · Module 3: Capacity and Cost Estimation · part 20 of 107
Part 20 · Module 3: Capacity and Cost Estimation

Lesson 1: From daily users to requests in flight

21 min read·
Module 3 · Lesson 1 · About 45 minutes · Needs: Module 2 lesson 3 (the constraints Joel wrote down), school arithmetic, and enough Python to run a cell. No model calls and no key.

The situation

Week 3 at Northdesk. Joel's target for the second quarter of 2027 sits in TARGET: 100,000 daily Ask users, 10 questions each, 2,000 input and 500 output tokens, a 3-second response, a peak factor of 4. Helen will not look at an architecture until the numbers are on one page. Funmi asks the first question: how many requests arrive in the busiest second, and how many are in progress at once?

This lesson answers with the traffic part of the roadmap's worked estimate; lessons 2, 4 and 5 do tokens, cost and latency.

THE TARGET Joel's target, Q2 2027 100,000 daily Ask users 10 questions each 4 peak factor 2,000 / 500 tokens in / out 3 s response target 60,000 Cases runs a day this lesson lessons 2 to 5 and the exercise
Joel's target holds six numbers; this lesson uses the three traffic ones (blue, Northdesk's system) and leaves tokens and time (grey) to later lessons.

The rest of this course is yours to keep

This course is bought on its own, once, and stays readable afterwards, including the parts added to it later.

Or open everything with a plan