This time to share a student's Visa 2026 New Grad OA experience. The overall feeling is almost exactly the same as TikTok's question bank (confirmed to be the same Codesignal question bank), the difference is only that the time has been shortened from TikTok's 70 minutes to Visa's 60 minutes (some batch even 45 minutes), but the content and logic of the questions are almost the same.
This student had practiced the TikTok version before, so he went straight to a full AC in 30 minutes, and for those who have received OA invitations but aren't quite sure, this is a complete reference.
OA Basic Information
- Platform: Codesignal
- quantity of questions: 4 Programming Questions
- timeframe: 70 minutes officially (only 30 minutes in real life)
- Difficulty distribution: 2 Easy + 2 Medium
- test languagePython / Java / C++, Python is the fastest!
This set of questions is not heavy on algorithmic fundamentals, but requires good code implementation speed and boundary awareness.
Question 1: Rating Calculation
Problem.
You are given an integer array diffs. Start with an initial rating of 1500. For each element in diffs, update your current rating by adding the difference. Track and return the highest rating you ever reach.
Example.
diffs = [100, -50, 200]
Output: 1650
Idea.
Simple simulation - keep two variables.
curr = 1500best = 1500
Iterate throughdiffs, updatecurr += diff, and keepbest = max(best, curr)
Complexity. O(n)
Question 2: E-Scooter Ride Simulation
Problem.
You are walking along a street with several scooters placed at different positions. Starting from position start, you need to reach end.
At each step.
- Find the nearest scooter to your right that you haven't used.
- Walk there.
- Ride 10 units forward (or stop at
endif less than 10 units away). - Repeat until you reach the end.
Return the total riding distance.
Idea.
Use a while loop to simulate the movement process.
At each iteration, locate the next scooter, move to it, ride, and update your position.
Complexity. O(n²) (acceptable due to small input size).
Question 3: Distribution Centers
Problem.
You have several distribution centers, each with a limited capacity. You receive a series of package logs - each package is assigned to the first available center with remaining capacity. You receive a series of package logs - each package is assigned to the first available center with remaining capacity.
When all centers are full, reset all non-closed Some centers may be permanently closed and should never reset.
Idea.
Maintain two arrays.
Capacity[i]for remaining capacityclosed[i]as boolean flags
For each log entry, assign the package to the first available center.
If all centers are exhausted, reset non-closed centers.
Implementation-heavy but straightforward.
Question 4: Similar Peaks
Problem.
Given an array heights and an integer viewingGap, find all pairs (i,j) Such that |i - j| >= viewingGap, and return the minimum absolute difference |heights[i] - heights[j]|.
Example.
heights = [1, 5, 2, 8], viewingGap = 2
Output: 3
Idea.
Brute-force all valid pairs using double loop, update the minimum absolute difference.
For small inputs, O(n²) is sufficient.
Summary and recommendations
On the whole, the question types of this Visa OA are almost identical to those of TikTok, with medium difficulty, and the examination points are focused on:
- Speed of implementation and code specification: Each question has a fixed logic and there are no complex algorithms.
- simulation capabilityThe questions, especially questions 2 and 3, are purely process simulation questions.
- Border control: e.g., endpoint judgments for scooter problems, reset conditions for distribution centers.
If you've already swiped TikTok OA, you can just seamlessly connect to Visa.
Wrap-up
The biggest problem with these Codesignal sets is that there are a lot of questions, time is tight, and it is easy to get stuck at the boundaries.
Many students get stuck on "analog logic" questions, such as the scooter question that is off or the capacity is not reset.
So if you don't want to risk figuring out the boundaries on your own, worrying about not having enough time.Programhelp Team Also available on a permanent basis:
No Trace Remote OA Ancillary Services:
The entire process is assisted remotely online through the security tool ToDesk, with no trace of detection on the system;
Adapts to HackerRank, Codesignal, Codility and other platforms;
Whether it's Visa, TikTok, Amazon, Capital One, we have a full library of questions and response strategies.
