Wipro Data Analyst Interview Questions
A data analyst at an IT services company often works on a client's operational data: support tickets, service targets, monthly reports and records from systems that were never built to agree with each other. Prepare to explain how you check data you did not create, describe queries in plain words, turn a hand-built report into a dependable one and tell a client something their numbers did not show. What to show here: numbers a client can trust in their regular service reviews. Each question below has what the interviewer is checking, an answer frame and a sample spoken answer to adapt.
How Wipro hires, in its own words, and its questions for every job →
1. Tell me about a time your numbers disagreed with what a client or manager believed to be true. How did you handle the conversation?
- Describe the belief and what your numbers showed.
- Explain how you checked your own work first.
- Describe how you presented it.
- Share what was decided.
“A manager I worked with was sure that most support tickets came from one old application. My count showed a different, newer system created more. Before saying anything, I checked my query, the date range and whether tickets had been moved between categories. The numbers held. I showed him the count along with a sample of real tickets, so he could read them himself instead of taking my word. I also explained why the old system felt worse: its tickets were louder, not more frequent. We agreed to your result. I learned to bring examples along with totals.”
Weak answers either drop their finding when challenged or announce it without checking their own query first.
2. Describe a time you found a data quality problem in a source system owned by someone else. How did you get it fixed?
- Describe the problem and how you noticed it.
- Explain how you measured its size.
- Describe how you raised it with the owner.
- Share the fix and how you guard against it now.
“While building a monthly report, I saw that many tickets had a closing time earlier than their opening time. I measured it: it affected tickets from one office, and only those closed by a certain tool. I could have filtered them out, but then the problem would stay. I sent the system's owner a short note with ten example tickets, the count and my guess that the tool used a different time zone. They confirmed it and corrected the tool. In my report I added a check that flags such rows. The result was your result.”
Weak answers silently delete the bad rows and tell nobody, so the source stays broken.
3. How would you measure the health of a support service that a team runs for a client? Which numbers would you put on the first page?
- State what a healthy service means for the client's users.
- Choose measures for speed, quality and volume.
- Add a balancing measure for each one that can be gamed.
- Show trends and splits, not single numbers.
“A healthy service means users get working answers quickly and do not need to come back. So I would show time to first response and time to resolution against the agreed targets, split by priority. To balance speed, I would show the share of tickets reopened and the user rating after closure, because closing fast is easy if quality does not matter. For volume, I would show tickets created against tickets closed, and the age of the backlog. Each measure would appear as a trend over several months, with the top categories listed, so the client sees causes as well.”
Weak answers show only the share of tickets closed within target, which says nothing about quality or repeat problems.
4. You have a table of support tickets with a priority, an opened time and a first response time. Talk me through a query that shows, for each month and priority, how many tickets missed their response target.
- Join tickets to a small table of target times per priority.
- Work out the response gap for each ticket.
- Flag breaches, including tickets with no response yet.
- Group by month and priority and count.
“I would keep the targets in their own small table, one row per priority, and join tickets to it on priority. For each ticket I work out the gap between the opened time and the first response time. A CASE expression marks a breach when the gap is larger than the target. Tickets with no response yet need care: if the target has already passed, they count as breached too. Then I GROUP BY the month of the opened time and the priority, counting all tickets and summing the breach flag. I would also ask whether the clock should stop outside business hours.”
Weak answers hard-code the targets in the query and forget tickets that never received a response.
5. A ticket can be reopened several times, and each change of status is a row in a history table. How would you find the tickets that were reopened more than once, and the team that closed them first?
- Filter history rows to the reopen events and count them per ticket.
- Keep tickets with a count above one.
- Rank close events per ticket by time and take the first.
- Join the two results and check a few by hand.
“From the history table I would first take only the rows where the status changed to reopened, group by ticket and keep those with a count above one using HAVING. Separately, I take the rows where the status changed to closed and number them within each ticket by time with a window function, keeping the first. That row gives the team that closed it first. Joining the two on the ticket id gives the list. Before trusting it, I would read the full history of a few tickets by hand, because status names are often used loosely in real systems.”
Weak answers use the latest status only and cannot explain how to get the first close event per ticket.
6. A monthly report for a client is built by hand from three spreadsheets and takes two days, with a different mistake each month. How would you make it dependable?
- Write down each manual step and where each figure comes from.
- Load the sources with a script and fixed rules.
- Add checks that stop the run when inputs look wrong.
- Run old and new side by side before switching.
“I would begin by sitting with the person who builds it and writing down every step, including the small fixes they do from habit. Then I would replace the copying with a script that reads the three sources and applies the same rules every time. Before any output, it runs checks: are all expected columns present, are totals within a normal range, is any month missing. If a check fails, it stops and says why. For two or three months I would produce both versions and compare them line by line. The layout the client knows stays the same.”
Weak answers rebuild the report in a new tool with a new layout and skip the side-by-side comparison.
7. On the morning of a service review with the client, the dashboard shows that resolved tickets fell sharply last week. You have an hour. What do you do?
- Check whether the data load finished and is complete.
- Compare with the source system directly.
- Split the drop by team, category and day.
- Tell the meeting what you know, what you do not and when you will.
“My first check is the data itself: did last night's load finish, and does the row count look normal? A sharp drop is often a missing file, not a real change. I would compare the dashboard with a direct count in the ticket system. If the data is complete, I split the drop by team and by day to see where it sits, such as one team on leave or a holiday. With the hour nearly gone, I would brief my lead plainly: this is confirmed, this is likely, this I will know by tomorrow. I would not guess in front of the client.”
Weak answers present the drop as real without checking the load, or hide the chart from the meeting.
8. Why do you want to work as a data analyst at Wipro, on data that comes from clients' day-to-day operations?
- Say what interests you about operational data.
- Connect it to client work in an IT services company.
- Give an example of a finding that changed a routine.
- Say what you want to learn.
“I like analysis that changes how people work on Monday morning. Operational data, like tickets and process times, is not glamorous, but a good finding removes a delay that hundreds of people feel. Wipro delivers IT services to client companies, so I would see that kind of data from different industries and learn what good looks like in each. In my last project, one chart about reopened tickets led a team to change its checklist. I want more of that. I also want to get stronger at building reports that run on their own and that a client can trust.”
Weak answers speak only about machine learning ambitions and show little interest in everyday reporting.
No figures on this page: pay depends on the level, the city and the offer in front of you. When the offer comes, the free salary negotiation script helps with that part.
The questions above are the prep. On the real call, ClapAssist helps with the ones you did not practice.
ClapAssist is an AI interview assistant for Mac and Windows. It listens to the interview on your computer and shows you what to say, in short lines you can read while you talk. Your live interview audio and screen are never stored. Your resume and notes are saved to your account so the app fills them in on any computer. It is hidden from screen share on every plan, including Free; only you can see it.