Mock interview scorecards and revision plan
Technical/concept scorecard (1–4 each)
| Dimension | 1 | 4 |
|---|---|---|
| Correctness | Material misconceptions | Precise guarantees/limits |
| Clarity | Unstructured jargon | Plain definition, contrast, scenario |
| Android depth | Generic software answer | Lifecycle/process/device/version implications |
| Practicality | Happy-path theory | Failure/corners/testing/operations |
| Trade-offs | “Best practice” | Contextual choice and rejected alternatives |
Coding scorecard
Contract clarification, cases, readable/idiomatic solution, correctness, complexity, cancellation/concurrency where applicable, tests and explanation. Do not reward cleverness that obscures behavior.
System-design scorecard
Requirements/non-goals; invariants/authority; high-level decomposition; read/write flow; local data/offline/sync; security/privacy; failure/recovery; tests/observability/release; scale/evolution; communication/prioritization. Score 1–4 and record one evidence quote, not a vague impression.
Leadership scorecard
Ownership, stakeholder understanding, decision quality, influence/conflict, execution, measured result, learning and growth of others. Watch for hero narratives that leave no healthier system/team.
Four-week revision loop
Week 1 — foundations
Kotlin/coroutines; lifecycle/process/components/Binder; daily code and 2-minute concept recordings.
Week 2 — Android delivery
Compose/Views; data/network/offline; security/performance/testing/release; one debugging drill daily.
Week 3 — architecture
Patterns/modularization/DI; ADR/trade-offs; three 30-minute designs; review weak chapters from score evidence.
Week 4 — simulation
Two coding, two Android deep dives, two system designs and two behavioral mocks; refine project walkthrough/resume; taper into recall/checklists.
Use spaced retrieval: attempt before reading, explain aloud, compare to chapter, correct a short error log, revisit after 1/3/7/14 days. Measure rubric trends, not hours watched.
Final readiness checklist
- Can explain core topics plainly before details.
- Can write/test Kotlin under a time limit and narrate trade-offs.
- Can debug from evidence without random-fix lists.
- Can design one read/write including offline, security and operations.
- Has 6–10 truthful leadership stories with metrics/learning.
- Knows current project architecture and personal contribution deeply.
- Asks role/team/quality/on-call/decision questions of interviewers.
- Has sleep, equipment, environment and clarification strategy ready.
Exercises
Run a complete loop, have a peer score it, select only the lowest two dimensions for the next session, and repeat until evidence reaches consistent 3–4 rather than rereading everything.