2 min lesson
Live authentic-usage drill
Walk through "Live authentic-usage drill" in order, then name the proof that tells you it worked.
Step 1 of 2
Live authentic-usage drillPractice out loud
- 1Pick a real task. Something you actually shipped with Cursor this week, with the repo, the surface (Tab, Agent, chat) and what you were trying to do.
- 2Be specific. Where it nailed the context, where it lost the thread, what you accepted and what you rewrote by hand.
- 3Give an honest critique. Name one real limitation you've hit - a latency annoyance, a context miss, a place a competitor does something better. Earnest praise with no critique reads as superficial.
- 4Tie it to the role. Connect a latency or reliability frustration you felt to the inference path you'd be working on.
"Why Cursor, why inference, why now: I use Cursor every day and the thing I keep noticing is that Tab feels instant - that sub-100ms budget is a genuine systems achievement and it's the part of the product I'd want to own. Inference routing is where reliability, cost and latency all collide, which is the kind of gray-area systems work I'm best at. And the field is moving fast enough that the platform decisions made now compound, so this is the moment I want to be doing it."
Keep the "why Cursor" under 90 seconds and end it on the product, not on yourself. "...and that's why I want to own the inference path" lands harder than a paragraph about your career arc. Then stop talking and let them follow up.
Learn more
Advanced table
Self-rubric
Self-rubricGrade your delivery
- Dimension
- Specific
- Pass
- Named systems, repos, surfaces, real numbers
- Red flag
- Generic "we improved performance"
- Dimension
- Quantified
- Pass
- p99 / error rate / cost with before-and-after
- Red flag
- Impact stated as adjectives
- Dimension
- End-to-end ownership
- Pass
- You carried it from design through prod and on-call
- Red flag
- Handed it off after the PR merged
- Dimension
- Honest
- Pass
- A real failure and a real product critique
- Red flag
- Polished only; no genuine miss anywhere
- Dimension
- Authentic usage
- Pass
- Specific daily-use detail and a grounded opinion vs competitors
- Red flag
- Marketing-page familiarity
| Dimension | Pass | Red flag |
|---|---|---|
| Specific | Named systems, repos, surfaces, real numbers | Generic "we improved performance" |
| Quantified | p99 / error rate / cost with before-and-after | Impact stated as adjectives |
| End-to-end ownership | You carried it from design through prod and on-call | Handed it off after the PR merged |
| Honest | A real failure and a real product critique | Polished only; no genuine miss anywhere |
| Authentic usage | Specific daily-use detail and a grounded opinion vs competitors | Marketing-page familiarity |
Bring three sharp questions for the team: how they balance cost vs latency when the two fight, what their on-call and SLO culture looks like for the request path and what onboarding-to-first-prod-ship actually feels like at their pace. Good questions are part of the signal.