The 90-Day Coding Interview Prep Plan (For Engineers With a Full-Time Job)
A realistic coding interview preparation schedule: DSA blocks, mock interviews, review cadence, and how to avoid burnout while working full time.
Most coding interview preparation plans assume you can grind eight hours a day. If you have a job, dependents, or both, you need a 90-day rhythm that compounds without wrecking your life. This plan balances problem practice, spaced review, and voice-first mock interviews so you train the same skills hiring loops actually measure—not only silent AC submits.
If you have not read our guide on how to think out loud in interviews, skim it first; this schedule assumes you are building communication alongside algorithms.
What “done” looks like after 90 days
- You can read a new medium problem, propose two approaches, and implement one in 25–35 minutes while narrating.
- You have 2–3 voice or human mocks per week in the last month (see AI mock interviews vs human for how to mix them).
- You maintain a review deck of ~40 problems you can re-solve cold.
- You can state time and space complexity without freezing—our time complexity guide pairs well with weeks 5–8.
Phase 1 (Days 1–30): Patterns, not trophies
Goal: Cover core patterns once; prioritize understanding over streaks.
| Week | Focus | Example topics |
|---|---|---|
| 1–2 | Arrays, hash maps, two pointers | Pair with Two Sum deep dive |
| 3 | Stacks, queues, monotonic stack | Parentheses, daily temperatures |
| 4 | Linked lists, fast/slow pointer | Cycle detection, merge lists |
Cadence (weeknights): 45–60 minutes deep work + 15 minutes spoken recap (“what pattern was this?”).
Weekend: One 2-hour block + one timed problem with a 35-minute cap.
Avoid: Chasing “number of problems solved” as your only metric. One explained problem beats three copied solutions.
Phase 2 (Days 31–60): Medium density + weak tags
Goal: Move from “I’ve seen this” to “I can execute under mild pressure.”
- Tag every solve:
two-pointer,binary-search-on-answer,BFS,topo, etc. - Each week, pick your worst two tags and schedule six problems there.
- Add one hard every weekend—not for ego, to stretch proof and resilience.
Introduce weekly complexity drills: after each session, write one paragraph explaining Big-O for that day’s hardest problem (ties to time complexity cheat sheet).
Phase 3 (Days 61–90): Interview fidelity
Goal: Match format, not just content.
- Two full mock interviews per week (voice AI or strong human). You want interruptions, hints, and time boxes—see what FAANG interviewers score so you know what signals you are optimizing.
- Alternate company-style sessions: one “strict bar,” one “warm coach.”
- Cut new problem intake to ~40% of time; 60% is review, weak tags, and communication.
Weekly template (copy-paste)
| Day | Activity |
|---|---|
| Mon | 2 mediums + verbal walkthrough |
| Tue | 1 medium + 1 review (blind) |
| Wed | Weak tag × 2 |
| Thu | 1 hard attempt (45 min cap) + post-mortem |
| Fri | Light review or rest |
| Sat | Mock interview + 1 medium cool-down |
| Sun | Review deck maintenance (add/remove cards) |
Adjust days to your energy; consistency beats hero days.
Burnout guardrails
- Cap daily new problems at three on work nights.
- If you miss a week, shrink scope—do not “double up” and quit.
- Sleep and exercise are part of prep; interviewers read cognitive bandwidth in how you recover from mistakes.
FAQ
How many LeetCode problems is enough?
There is no magic number. For most mid-level roles, 120–180 quality solves with spaced review outperforms 400 shallow ones. Quality means: you can re-derive the approach and complexity.
Should I use a premade list?
Lists help structure; your weak tags should override generic order after week four.
Where does TechInView fit?
TechInView is built for voice DSA mocks with live coding—ideal in phase 3 when you need interview fidelity. Start a session when you are ready to practice speaking and coding together, not only typing.
Summary: A sustainable coding interview preparation plan runs in three 30-day phases—patterns, density, then mocks—with explicit time for communication and review. Treat the last month like the real interview: timed, vocal, and honest about weak tags.