EBoard 08: Behavioral interviews and more
Approximate overview
- Administrivia
- Visitor talk(s)
- Questions
- Reflections on practice (or non-practice) behavioral interviews.
- Brainstorming strategies (if time)
Administrivia
Visitor Talk
- Class of 200x.
- Works in Seattle at a consulting firm: Web apps, mobile apps, AI projects
- 11 years at the current place
- Path …
- Worked in Iowa on e-government stuff
- Lots of PHP, taking paper processes, turning them into online processes
- Campsites, license renewal, and more
printwhatyoulike.com (still used 15 years or so later)- Freelance work (plus travel)
- Consulting can be very flexible
- Moved out to Seattle
- Has grown with the company
- The tech industry is in a rough patch
- Survived the round(s) of layoffs
- However, it’s been a rough year
- Investment money is dried up, except for AI; since they consult
for startups
Questions
———
What is consulting?
- If you work at some place like Amazon, you work on their product.
- They work for assorted clients who say things like “We want a Web app.”
Clients come with a budget. This is what we can build for you …
- From a few months to a few years.
- Opportunity to learn new things.
- Nice for variety.
At a consulting firm, do people stay within a particular area?
- When mobile apps were hot, the mobile app developers not only developed
mobile apps, but trained others to do so.
- If you get hired on a multi-year project, you’ll probably work on that
for a few years.
When you have to switch technologies, how do you get up to speed?
- It depends on the technology.
- You learn on the job. Often, if you have other people on the job who
know it, you learn from them. Otherwise, you teach yourself.
- Example: 3D modeling; had to figure it out from scratch. Looking forward
to showing it off.
How are projects assigned or distributed amongst people?
- When they had twenty SWDs (software developers), they had a few
folks per project. Now they try to get at least two, or at least
one plus who is available to help.
- Also try to follow people’s preferences.
- And there’s some swapping.
What’s the distribution of people?
- There used to be designers. (When they were healthier.) Now they
contract out for design.
- Engagement managers: Project managers + Product managers + Client
relationships.
- Research and strategy. They help clients figure out what they want.
Writing documents. Rethinking math education to make it more equitable.
However, doesn’t lead into product.
How do you get clients?
- Most of their work now comes from referrals.
- Their CEO knows a lot of people. Good at meeting people.
What does doing AI mean these days?
- 80% of it is normal app development. Not really model development.
E.g., Find new ways to connect people based on how they use the web.
- You can do a lot with ChatGPT.
- E.g., Resume analyzers plus job postings. Interesting that it
doesn’t always the same score.
Backing up, tell us about the 3D model web site?
- Used 3JS.
- Someone else built the 3D model; the alum helped them learn the
JavaScript.
- Figuring out the lighting and materials was interesting. Lots
of tweaking and learning.
Ethics of digital nomad lifestyle
- Most people don’t do it long term
- Those air-bnbs price out locals
- Support the local economy! (We hope that helps.)
- Some places are trying to attract digital nomads; go to places
with struggling economies
What is your work week like?
- 3/4 time working on projects
- 1/4 time directing stuff
- Manages two people (meets once per week)
- Participates in company planning
- Working on learning the whole AI thing; how do we get people
familiar with these tools so that they can be put on projects?
- Clients don’t really know much about AI, so preparing demos
is good. Getting demos put together.
- Learning and documenting to help others.
- Tons of database stuff to get things working smoothly.
- Helping with project
What does project development look like?
- Most weekdays involve standups with clients to discuss where things
are going.
- A bit of pairing when people are working on something difficult;
pairing helps dig yourself out.
- A lot of coding on your own.
- Sometimes investigating for estimation.
- Do you pull in component libraries or build them yourselves?
- A lot of time talking.
- There is no typical day.
- Another thing … upgrading
What was your college journey?
- Came in and wanted to do CS. Really liked programming in high school.
- Started in 153 (a combination of the first two courses).
- Still likes Scheme. Yay! Sam is happy.
- Studied abroad in Ecuador.
- Did a MAP with Prof. Kuiper on modeling bison/cattle and prairies.
- Took the main set of CS classes, although not software development.
- Most of the languages they’ve used they learned afterwards.
PHP, Ruby/Rails, Swift, JavaScript, Python (doesn’t like it).
Question for us … what does the intro curriculum look like?
Question for us … what does 324 look like?
- Basic R, basic agile, …
- Note for us: Things will change; learn how to learn!
- Also ethics and product design
How did you document your journey with AI?
- Work log: Write down what you did. Put down error messages.
- It’s all in a GitHub repository.
- Onboarding helps. More documentation helps people join projects.
- Maybe a ‘blog post (eventually).
- Using the work log is somewhat ad hoc.
- Sometimes people tag stuff.
How did you learn AI?
- Playing: It’s interesting to see what ChatGPT comes up with.
- Amazon AI hiring fiasco; think about your data
- Lots of AI ethical issues
Tips on how to be successful interviewees …
- Initial screening to make sure that we’re not wasting each others’ time.
The hiring process is time consuming.
- They try to do a process that mimics what work day is like.
- A coding session.
- A pairing session. “Make a feature change.” Much more of a
collaborative session.
- The last time they were on the market, a lot of interviews were mostly
“brain teasers”.
- Are they a good technique for finding employees? Possibly not.
- Unfortunately, things have switched from “candidates have power” to
“companies have power”.
- Goal of checking fit.
- “Ask questions!”
- Figure out what you want in a job; does the company have what you’re
looking for? “What do you like about your job.”
What makes a person a good fit? (at least at your company)
- Collaborative. For example, the interview sees how you work with a
designer and with a peer.
- Some level of technical skill. A coding exercise.
- Communication. How do you deal with clients, particularly non-technical
clients?
- Wants to learn. Can learn quickly.
How do you assess learning?
- “What’s the most recent technology that you experimented with?”
- Plus, you learn new things asking that.
How do you keep yourself organized?
- The work logs have been helpful.
- Prioritize, prioritize, prioritize.
- Keep notes on hand. Before leaving one task, do a brain dump so
that you can come back.
How do you ethically negotiate with your clients?
- Some clients are more technical than others.
- Try to be as transparent as possible.
- It’s up to the client to decide what they want to do. The
consultant provides the recommendations (and choices).
How do you estimate?
- That’s a tough one. Estimation is really hard. It’s necessary
to pad.
- “There’s a range. If everything goes really well, it will take
this long. However, not everything will go well, so it may
take as much as this long.”
- Checkpoints along the way are a good way to go. It’s not an
“all at once” process.
What did you do at Grinnell that helped prepare you?
- Learning how to learn.
- Explore your interests!
- Knowing more than one language helps, at least for some projects.
- Communication. Working on team projects helps. There are very
few jobs in which you work by yourself.
AI is currently big. How long will it stay big/a trend?
- AI is around for awhile.
- Worries about the ethics.
- Worries about automating people’s jobs away.
The next trend?
- Web 3.0? (No clients in that direction.)
- Quantum computing?
- The return of PHP.
Does anyone still use PHP?
- “Half of the Internet runs on PHP.”
Who maintains the software?
- “We are a boutique consultant. We’re pretty expensive.”
- In the ideal world, the client hires someone to maintain.
- But they get sucked into some maintenance contracts.
- And those usually need quick turnaround.
- “Software development is not a one-time process.”
What question should they have asked?
- What should you look for at a company? Leadership matters. Culture matters.
Values matter.
- Of course, you may just need a job.
- Work/life balance may be important to you.
- What do you want.
Behavioral Interviews
Nope.
Brainstorming Strategies
a.k.a. Strategies for brainstorming
You’ve been asked a technical question. You’ve established knowledge of
the problem by asking questions. Now what?
That is, you have an algorithmic problem. How do you get started coming
up with solutions?
No time.