Sorting Questions at Infosys: What to Expect

Sorting questions appear in nearly one-fifth of Infosys's technical assessment problems (27 out of 158). This frequency reflects a core truth: sorting is rarely the end goal, but it is a fundamental step that enables efficient solutions to more complex problems. Infosys uses these questions to evaluate a candidate's ability to recognize when data organization is required, implement efficient algorithms, and apply logical thinking to real-world data processing scenarios—skills directly applicable to the large-scale system development and maintenance work the company undertakes.

What to Expect — Types of Problems

You will not be asked to simply implement a standard sorting algorithm from scratch. Instead, problems integrate sorting as a critical step. Expect these categories:

1. Sorting with Custom Comparisons: The most common type. You'll sort objects, strings, or arrays based on non-standard rules (e.g., sort numbers by the number of factors, sort strings by length then alphabetically).
2. Efficiency-Driven Sorting: Problems where choosing the right sorting approach (like using Counting Sort for a limited range of integers) is key to meeting time/space constraints.
3. Sorting as a Pre-processing Step: The sorting operation itself is straightforward, but it transforms the data to simplify the core logic, such as finding a minimum difference, identifying duplicates, or merging intervals.
4. Hybrid Problems: These combine sorting with other core concepts, like searching (binary search on a sorted result), two-pointer techniques, or greedy algorithms.

How to Prepare — Study Tips with One Code Example

Master the theory behind O(n log n) comparison sorts like Merge Sort and QuickSort, but focus your practice on applying the built-in sort functions intelligently. Your primary skill is learning to define the correct sort key or comparison function.

Key Tip: For any problem, ask: "Would sorting the input array make the solution easier or more efficient?" If the answer is yes, determine exactly how the data should be ordered.

Consider this common pattern: Sorting based on multiple criteria. Many problems require a primary sort key and a secondary key to break ties.

# Sort a list of employees: primary by department (ascending),
# secondary by salary (descending)
employees = [
    {'name': 'Alice', 'dept': 'IT', 'salary': 90000},
    {'name': 'Bob', 'dept': 'HR', 'salary': 75000},
    {'name': 'Charlie', 'dept': 'IT', 'salary': 85000}
]

# Using a lambda to return a tuple (primary_key, secondary_key)
employees.sort(key=lambda x: (x['dept'], -x['salary']))
print(employees)
# Output: [{'name': 'Bob', 'dept': 'HR', 'salary': 75000},
#          {'name': 'Alice', 'dept': 'IT', 'salary': 90000},
#          {'name': 'Charlie', 'dept': 'IT', 'salary': 85000}]

Recommended Practice Order

1. Start with basic custom sorting using your language's built-in tools (e.g., sort(key=) in Python, Array.sort() with a comparator in JS/Java).
2. Practice problems where sorting is the primary enabler: finding min/max differences, checking for anagrams, or merging overlapping intervals.
3. Tackle hybrid problems that combine sorting with binary search or the two-pointer technique.
4. Finally, understand the edge cases where a non-comparison sort (like Counting Sort) would be optimal, as Infosys problems sometimes test for efficiency awareness.

Practice Sorting at Infosys