EBoard 04: Thinking aloud

Approximate overview

• Administrivia
• Visitor presentation
• Sample tech interview one
• Sample tech interview two
• Debrief
• Steps for success, revisited (if time)

Administrivia

• Welcome to our visitor.
  • Our visitor may be doing different questions than I sent out. (Don’t worry, the voluntolds did not get the questions.)
• Next week: Halle Remash returns to campus (virtually).
• Expect to see the visitor schedule evolve over the next few weeks.
• I may steal a week for a return to “What skills should you practice for your technical interview?”

Visitor Presentation

Sam may attempt live notes.

Background

• Started at Grinnell in 2000. [Wow, that’s a long time ago.]
• Planned to do math. Took CS on a lark. Got stuck with SamR.
  1. Loved it.
• [Snarky comments about other majors.]
• Networked a lot working with people in Noyce late at night.
• Didn’t do any internships while at Grinnell. Did summer research in Mathematics in Summer 2003.
• Had no plan when graduating in 2004.
• Faculty member forwarded along an email. Moved out to California. Worked for O’Reilly Media.
• Decided to go to grad school. Went to Vanderbilt. Did a Ph.D. In Nashville.
• In 2011, started looking for jobs. Applied to Amazon (job offer), Google (rejected), NSA (hired).
• Worked for the government for two years. [Learned how to spy on US Citizens.]
• Decided to interview with Google again. Read Cracking the Coding Interview this time.
• “Interviewing is a skill. Like any other skill you can get good at it.”
  • It makes all the difference.
  • [Hmmm … we’ve heard that before.]
• Four years in Mountain View. Then got married and moved to Switzerland.
  • Eventually ended up as a manager.
• In March of 2023, laid off. (The whole division was laid off.)
  • Prepared for managerial interviews. Came up with 50 or so questions.
    • “Tell us about a time you had difficulty with a team member”
  • Lots of coding practice too. Decided to interview in Python.
  • 30 minutes per day practicing. Doing lots of other things.
  • Got an offer in April. Stayed in Switzerland until June.
• Currently a senior engineering manager.
  • Manages seven people.
  • Does a lot of interviewing / screens.
• Has kids and is about to buy a house. Yay real life!

Questions

How did you structure your preparation?

Remember: You’ll be presented with a problem. You should never jump into the solution right away. Ask questions. It shows the interviewer that you care about the problem. It also helps ensure that you are thinking about the same problem.

Sample problem: “You have to parse a file and extract all the phone numbers.”

Sample questions:

• “What kind of file is it?”
• “Will it fit in memory?”
• “What do phone numbers look like?”
• “Do I have to validate that they are real phone numbers?”

Make sure to think aloud while you are developing the solution in the editor.

Be creative in your test cases. Think about edge cases.

In preparing for this set of interviews, wrote a bunch of small questions that helped them ensure that they knew the core aspects of Python.

“In real life, you can look things up.”

In technical interviews, there’s coding and algorithms. They were pretty sure that the understood the algorithm.

What was your thesis?

Robust routing in wireless mesh networks.

Never worked on networking again.

Some time during grad school, decided that academia wasn’t the right direction for them.

Cared about flexibility in place and work/life balance.

What did you do at Google?

For four years, worked on Google Analytics. They worked on a product that told advertisers when someone searched for a product immediately after an ad was shown.

Worked on the front-end.

In 2017, moved to a new team (and a new country).

Worked on Google shopping to speed up databases. His team provided data.

What is a workday like?

Once they became a manager …

“A lot of meetings.” “More than half of my time is in meetings.”

A lot of stand-up meetings with two-week sprints.

Meetings with other companies (e.g., third-party vendors that are doing software as a service).

Lots of hiring meetings with candidates.

Responding to questions on Slack.

Forwarding tickts to the next person.

Doing forward planning. Small company (500-people) planning is very different than Google-size planning. Small companies have to be ready to pivot.

What was working for the NSA like?

The interview was mostly background checking (they got a top-secret background clearance; it took 4-5 months).

Rotated between several (three) different offices.

First one: Processing data. Wrote a lot of Perl scripts and Python scripts.

Second one: Cloud analytics fusion (CAFe). Big data store.

Third one: “Setting up some database stuff.”

How is government different than academia or industry?

There’s more stability in the job. “The government doesn’t really do layoffs.”

On the other hand, there are fewer perks and smaller salaries.

Google has nap pods, free food, buses, and more.

Congress doesn’t like government offices to spend money that way. NSA parking was awful. Before 7:30 a.m.: 1/4 mile to office. After 7:30 a.m: 1/2 mile to office. After 8:00 a.m., “Park on the other side of the interstate and wait for the shuttle.”

But you have a public interest mission rather than a “make money” mission.

Sample Tech Interview One

Sam is not recording the question or the work. Just a few comments.

There would normally be some introductions and then a short, partially ambiguous question.

Note: There are normally stages to a question. How would you add X to your code?

Sample Tech Interview Two

Always start with questions.

Work through some examples.

Make sure that you have time to do the algorithm.

Pause and think through the algorithm.

Think through your data structures.

Come up with some test cases/unit tests.

Debrief

Notes

• Ask for help!
• Think about other directions!
• Clarifying questions are nice way to start things off.
• Don’t start coding too early.
• Try to predict what kinds of questions the interviewer might ask and then ask them of the interviewer.
• Think about tradeoffs (maintainability and readability)
  • Interviewers expect you to discuss them.
• People care about process as much as the solutions.

Sam’s questions

• Good questions:
  • What is the format of the input?
  • Should I handle things that are outside.
  • What kind of security do we want?
  • Do I have access to these data.
  • What is the return type?
  • Are there naming conventions I should follow?
  • What should I do in these exceptional situations?
• They were good at responding to the interviewer’s questions and in asking their own; there was a conversation.
• They weren’t rushing.
• Interesting example: One of the first mars landers crashed because of inconsistency in units throughout the program.
• Good at talking about where they are getting stumped.
• Know some strategies. E.g., “Convert it to a graph and use a graph algorithm”, “Use dynamic programming”, “Use a greedy algorithm”
  • What will the graph look like?
• Criticisms:
  • Don’t work with a language you don’t know (well).
  • More object thinking!
  • Make sure to discuss the algorithm before coding.
  • Know your data structures/ADTs/algorithms.

Tips

• Know your classes. There’s more than just algorithms and data structures in your solutions. Sometimes you get design problems.
• Learn how to use a mac keyboard.
• Find the right balance of when to jump into the code.
  • You may want to explicitly address that with the interviewer.
  • “Is this the right time to start writing some code?”
• It’s okay if you have a bad interview. Think about how you can give the interviewer positive signals. Keep the momentum going.
• Interviewers may feel like a blank wall; you will likely have to take initiative.

Steps for success, revisited

See the course schedule.