Interview Cheat Sheet

Quick reference for product-focused frontend interviews (e.g. 45-minute live coding). Focus on clarity, correctness, and calm reasoning over cleverness.

1) Core patterns (most questions)

Tree / nested data

Components, folders, JSON. Classic DFS.

function dfs(node) {
  visit(node)
  for (const child of node.children ?? []) {
    dfs(child)
  }
}

Time: O(n) — visit each node once. Space: O(h) stack + O(k) output.

Sliding window

Strings, tokens, event streams.

while (right < n) {
  add(right)
  while (valid()) {
    updateAnswer()
    remove(left)
    left++
  }
  right++
}

Time: O(n). Space: O(k) (distinct required items).

Hash maps / frequency counts

map[key] = (map[key] ?? 0) + 1

Time: O(n). Space: O(u) (unique keys).

Intervals / ranges

Sort by start, then sweep and merge.

Time: O(n log n). Space: O(n).

2) String problems — key distinctions

Literal substring → indexOf / simple scan
Any permutation → sliding window + counts
Min window containing → expand right, shrink left
Words instead of chars → tokenize, same window logic

3) Tokenization + normalization (product FE)

Tokenize original text first
Store raw, normalized, start, end per token
Match on normalized, slice original using indices

Normalization cost: O(T) total characters.

4) TypeScript live-coding tips

Prefer for…of over reduce under pressure
Use Record<string, number> for counts
Guard optional fields: children ?? []
Early returns for edge cases
Readable > clever

5) Interview pacing (45 minutes)

0–5 min: clarify + examples
5–10 min: approach + complexity
10–30 min: code correct solution
30–40 min: edge cases + cleanup
40–45 min: tradeoffs / alternatives

Say early: “I’ll start with a straightforward solution and optimize if needed.”

6) End-of-interview questions

What kinds of problems does this role solve most often?
How do frontend and platform teams collaborate?
What does success look like in the first 90 days?