Bloomberg 26NG Superday Coding Recap: Max-Depth Parentheses Substrings + 3 BQ Prompts

This was a Bloomberg 26NG Superday coding problem. The algorithm is straightforward, but communication quality is what really differentiates candidates.

---

Clarifications first

I would confirm two points up front:

1. Is the parentheses string always valid? Yes.
2. Can the content inside parentheses be empty? Yes, e.g. ().

These clarifications keep the implementation focused.

---

Improved interview wording (ready to say)

I scan left to right and maintain three states:

• depth: current nesting depth
• maxDepth: maximum depth seen so far
• curr: substring collected for the current nesting layer (excluding outer parentheses)

Rules while scanning:

1. On '(': Increase depth and start collecting for a deeper layer. Deeper layers are the only candidates for larger depth answers.

2. On ')': We are about to leave the current layer. If depth > maxDepth, clear previous answers, add curr, and update maxDepth. If depth == maxDepth, append curr. Then clear curr and decrement depth.

3. On normal characters: Append to curr.

At the end, return all substrings associated with maximum depth (format depends on the prompt).

---

Complexity

• Time: O(n)
• Space: O(n) (result + temporary substring)

---

Common pitfalls

• Updating depth and maxDepth in the wrong order
• Forgetting to clear old answers when a new deeper level appears
• Failing empty-content cases like ()
• Accidentally including outer parentheses in returned substrings

---

3 practical BQ prompts (for Bloomberg Superday)

BQ1: Tell me about a time you fixed a high-pressure issue quickly

Structure:

• Situation: critical bug/outage context
• Action: triage, isolate, patch, monitor
• Result: recovery time and prevention improvements

BQ2: Tell me about a disagreement with a teammate

Structure:

• Conflict: what exactly differed
• Action: align goals, define metrics, run a small validation
• Result: data-driven decision and team learning

BQ3: Tell me about a time you improved efficiency proactively

Structure:

• Bottleneck you identified
• Change you designed and shipped
• Measurable impact (time saved, error reduction, throughput)

---

Takeaway

This problem is essentially one-pass state tracking. In Bloomberg-style interviews, correctness matters, but clear state reasoning and edge-case communication matter just as much.

---

If you are preparing for Bloomberg / Google / Amazon interviews, feel free to reach out.

#bloomberg #superday #vo #newgrad #sde #interview

---

Further Reading (External Links)

• Bloomberg Careers
• Longest Valid Parentheses (LeetCode 32)
• Stack (Wikipedia)

---

Need real interview questions? Contact WeChat Coding0201: Get Questions.

Contact

Email: [email protected]
Telegram: @OAVOProxy