Interview question: what is the first thing you do?

dl324

Joined Mar 30, 2015
12,697
What sort of answers do you think they are looking for?
There's no canned answer.

Obviously you need to make sure you understand the high level requirements because you'll either be given a schedule or be asked for one. So you'll be asking questions to clarify your understanding.

You'll also need to understand what resources will be available and determine if they'll be adequate to complete the task in the specified amount of time.


When I worked on supporting layout verification for large microprocessor projects, I had to estimate how long it would take (for tapeout) and how much compute resource I'd need. It took many months to prepare mock data to see if additional compute resources needed to be purchased (which had to be forecasted at least a year in advance).

For one project, I forecasted that we would exceed the memory capacity of any 32 bit compute server we had or were planning to buy. I needed to acquire a 64 bit machine and migrate our infrastructure to the new hardware (and OS) by myself to create a test environment. The machine was loaned to me by the vendor. I had to get 64 bit versions of the programs from internal and external sources for evaluation and get hundreds of thousands of dollars budgeted for new compute servers.

I needed to compare results between the 32 and 64 bit solutions on "meaningful" sizes of data to verify that the new software gave equivalent results (I couldn't get the same versions for both 32 and 64 bit because they had diverged). Then I had to mock up a database that would be representative of our completed project (2 years in advance of the actual data being available). This was the largest design that any division at our company had contemplated because my project was the lead design on the latest process that the company was in the process of developing.

The worst part of the process was battling the naysayers who were saying I was crying wolf and that the current software and hardware platform would be sufficient. I knew my job better than anyone else on the project (and it turns out the whole company because one individual on my team (a self appointed personal antagonist as he put it in one meeting) solicited expert advice (because he had no expert knowledge of his own) from other parts of the company). In the end, I was right and they were all wrong. If they had managed to convince management that I was wrong (and believe me, they did try), the project would have been delayed by 6 months to a year. That would have cost the company billions of dollars.
 
Last edited:

MrSoftware

Joined Oct 29, 2013
2,011
OK it just hit me... clearly they were trying to figure out if you were management material. The answer they were looking for was, "can I outsource this task, or must it remain in-house?" .

https://www.cnn.com/2013/01/17/business/us-outsource-job-china/index.html

PS-> Though what this guy did was not right, he is still my hero in some small "Office Space" kind of way. Had he been in management he would have been given a fat bonus and a raise, but since he was in engineering he got fired. ;)
 

Thread Starter

bug13

Joined Feb 13, 2012
1,948
OK it just hit me... clearly they were trying to figure out if you were management material. The answer they were looking for was, "can I outsource this task, or must it remain in-house?" .

https://www.cnn.com/2013/01/17/business/us-outsource-job-china/index.html

PS-> Though what this guy did was not right, he is still my hero in some small "Office Space" kind of way. Had he been in management he would have been given a fat bonus and a raise, but since he was in engineering he got fired. ;)
Damn it, I didn’t know I was being interviewed for a management role. Well too late now, I should have known!
 

GetDeviceInfo

Joined Jun 7, 2009
1,814
I think it's probably a bit early for #2, but #1 is right on track.

I hate questions where you are having to guess what the person asking the question had in mind, but if they are any good they will keep an open mind and consider whether or not what you responded was reasonable regardless of what they were hoping for.

I would have said something along the lines of: The hardest part of solving any problem is usually understanding what the problem really is. So I would make sure that I had a firm understanding of the task, to include not just what the specific task was, but also, to the degree possible, what the bigger problem is that this task is a part of -- it could be that there is a completely different, but better, way of solving that bigger problem. Then I would make sure that I had a good understanding of what constitutes success in completing this task; how good is good enough? When does it have to be done by? What resources are available and what constraints must I live within? What aspects of it are the most important and which ones are less important? Who is the person that I can interface with that has the knowledge and authority to answer questions and make decisions about this task? Of course, the amount of time I spend gathering all of this information is going to depend on the scope and cost of the task.
This would be my approach. However, one must evaluate who is asking. An old saying in sales, no use wasting breath on a person who can’t make the buying decision. A micro manager could interpret this response as intimidating. I usually condense the whole thing by stating that I would ensure that my vision of success matches the companies vision of success. If they asked how I would do that, I’d get into all those details.
 

WBahn

Joined Mar 31, 2012
26,398
This would be my approach. However, one must evaluate who is asking. An old saying in sales, no use wasting breath on a person who can’t make the buying decision. A micro manager could interpret this response as intimidating. I usually condense the whole thing by stating that I would ensure that my vision of success matches the companies vision of success. If they asked how I would do that, I’d get into all those details.
No offense (seriously -- I'm not trying to offend), but if I asked a potential candidate a question and their response was that they would ensure that their vision of success matches the company's vision of success, I'd be looking for a way to end the interview as quickly as I could to move onto someone that wasn't going to just shower me with stock, meaningless suck-up platitudes. Maybe that approach does work with some hiring managers, but it sure wouldn't work with me and I'm pretty sure it wouldn't have worked with the vast majority of people I've worked with and for over the years.
 

cmartinez

Joined Jan 17, 2007
7,294
Hi guys

Just try to understand an interview question I got asked a while ago. The role is intermediate Embedded software engineer. One of the question I think I definitely got it wrong is:

And I answered:

But I think those are not the answers they are looking for. What sort of answers do you think they are looking for?

Thanks guys!
I think I'd go for something in the line of:
  1. Thoroughly understand the goal
  2. Learn all of the resources available at my disposal (tools, time, money, etc...)
  3. Learn how perfect a solution the employer/customer wants or needs (i.e. developing software can be a bottomless pit if one doesn't set limits)
I believe there's a distinction to be made between a problem and a goal.... sometimes the problem being presented distracts us from the goal itself.
 
Top