Zepto Practice: Operations Data and Clear Code

This page contains original exercises about operational data and code. It is not Zepto placement-paper material, a record of Zepto business metrics, or a current recruitment guide. For a live application, check the official Zepto careers page, the current role posting, and the written information attached to it. This guide does not claim a delivery promise, company scale, role availability, eligibility threshold, compensation figure, interview sequence, or selection outcome.

Model an inventory exercise as fictional data

Create a small fictional list of items with a name, available quantity, and reorder level. Write a rule that selects items requiring attention. Before writing code, decide whether an item exactly at its reorder level should be included and what happens if a quantity is missing.

Label the list clearly as an exercise. A number in a fictional problem is useful for testing logic, but it is not evidence about a real company's inventory, customers, or operations. Explain how you would change the design if data arrived late or a stock count needed a human correction.

Choose a queue rule deliberately

Suppose requests arrive with a priority and a creation time. Describe how you would select the next request, including a tie rule. Then describe a situation where strict priority might be unfair and a different rule could be needed. State the test cases that show whether your selection function follows the chosen rule.

This kind of answer demonstrates algorithmic thinking without pretending to know how any specific company dispatches work.

Investigate a mismatch between two records

A dashboard says an item is available while a downstream system says it is unavailable. List the evidence to gather: record identifiers, event times, source versions, and the transformation between systems. Explain how you would avoid making a broad repair before identifying the failing boundary.

If a correction is needed, preserve an audit note that says what changed, why it changed, and how the result was checked. A fix that cannot be explained later is difficult to trust.

Explain a performance trade-off

Take a lookup that becomes slow as a list grows. Compare scanning the list with building an index. Explain the extra memory cost, the need to update the index when data changes, and the observation that would justify the more complex approach. Do not promise a performance gain without an actual measurement.

The answer should be specific enough that another developer can turn it into a benchmark and challenge the conclusion.

Use live employer information only when acting

When you decide to apply to a role, open the current official posting and follow its own instructions. A previous listing, a social post, or a generic preparation guide cannot establish a live application rule. Use the named official channel for an important clarification.