1. David Robinson, Data Scientist
1.1. Believe in producing public artifacts
1.1.1. Sometimes benefits show up months or years down the line and lead to opportunities I never would have expected
1.1.2. Makes it easy to evaluate a job candidate: if I can look at some graphs created, how they explained the story, how they dug into data
1.1.3. Anything still on your computer, however complete, is worthless
1.1.3.1. Anything out in the world is much more valuable
1.1.3.2. I did a lot of things that I never published - kept feeling they weren't ready
1.1.3.3. Now I've forgotten what's in them, can't find them, and they haven't added anything to the world
1.1.3.4. If I'd written a couple blog posts, maybe made a really simple package, all of those would have added value
1.2. Data analysis should be a habit
1.2.1. If you see an opportunity to analyze data, take a quick look - see what you can find in just a few min
1.2.1.1. Decide on a set amount of time, do all the analyses you can, then publish it
1.2.1.2. One dataset => one post can create the habit for you
1.3. Advice to a junior data scientist
1.3.1. Don't stress yourself over keeping up with the cutting edge
1.3.2. be comfortable transforming and visualizing data
1.3.3. programming with a variety of packages
1.3.4. statistical techniques like hypothesis tests, classification and regression
1.3.4.1. understand these concepts and get good at applying them
1.3.4.2. do this before worrying about concepts at the cutting edge
2. Kristen Kehrer, DS instructor
2.1. Update your resume often
2.1.1. Esp if you've been working at the same place for a while!
2.2. Rank your resume high in terms of matching keywords
2.3. Optimize your resume for the job you WANT, not the job you have
2.4. What to show in your resume
2.4.1. You can solve problems
2.4.2. You can manage yourself
2.4.3. You can communicate well
2.4.4. You can achieve results
2.5. Start applying!
2.5.1. Don't keep taking online courses!
3. Brooke Watson, senior DS at ACLU
3.1. What to consider besides salary
3.1.1. lifestyle
3.1.1.1. Day-to-day way that the job interacts with your life
3.1.2. learning
3.1.2.1. Am I going to grow in this role?
3.1.3. values
3.1.3.1. Do this org and team work toward something that aligns with my values?
4. Hilary Parker, DS at Stitch Fix
4.1. Don't do a stream-of-consciousness notebook for analysis
4.1.1. You should flip that to "Here's the conclusion, and in the appendix, you see where I started"
4.1.2. What's going to be the easiest to produce quickly may not be the most readable
4.2. How to handle people asking for adjustments
4.2.1. They might be articulating unease
4.2.2. You should figure out what the person is actually trying to say
4.2.3. What's the root cause? Is that something that's appropriate for an analysis to address?
5. Sam Barrows, DS at Airbnb
5.1. A tool for working with stakeholders is to turn requests into dialogues
5.1.1. Interest-based negotiation
6. Amanda Casari, Engg Mgr at Google
6.1. When to start looking for a new job?
6.1.1. My role
6.1.1.1. I enjoy working on projects at their start
6.1.1.2. I also think about the stage of team cohesion
6.1.2. Where the product is in this product life cycle
6.1.3. Where the team is in forming vs conforming
6.2. Don't think you're irreplaceable
6.2.1. Nobody else is going to execute your job exactly like you do
6.2.2. But that doesn't mean that no one else can do it
6.3. Experienced people
6.3.1. Simply doing what's asked of you is no good
6.3.2. Larger kinds of impact and change
6.3.3. Solving problems
6.3.3.1. Scalable
6.3.3.2. Repeatable
6.3.3.3. Solves an org problem
7. Michelle Keim, head of DS and ML at Pluralsight
7.1. To prevent your project from failing, you really have to get down into the weeds and understand the use case of the problem
7.2. Know what success looks like at different points along the way!
7.2.1. Lets you validate early
8. Vicki Boykis
8.1. We're now at high-growth adoption across larger companies
8.2. Standardization of DS workflow toolsets
8.3. There is oversupply of new data scientists
8.3.1. Job market is extremely competitive and crowded
9. Robert Chang
9.1. When you're building out data capabilities, you have to build it layer upon layer
9.1.1. See Monica Rogati
9.2. what to look for when job-searching
9.2.1. When looking for DS job, focus on state of data infrastructure in the company
9.2.2. otherwise it will take months if not years to do any ML
9.2.3. then look for the people
9.2.3.1. experienced leader
9.2.3.1.1. knows how to build and maintain a good infra and workflow for DS to be productive
9.2.3.2. manager who is supportive of continuous learning
9.2.3.3. work with a tech lead or senior DS who is very hands-on
9.2.3.3.1. for your day-to-day work, that's the person who helps you the most
9.3. some companies look for unicorns
9.3.1. data wrangling skills with R or Python
9.3.2. building ETL pipelines
9.3.3. data engineering
9.3.4. experiment design
9.3.5. building models and putting them into production
10. Randy Au, UX researcher, Google
10.1. Small and large companies
10.1.1. 10: everyone does everything
10.1.2. 20: start getting 3-4 person teams
10.1.3. 80-100: existing teams don't scale, lot more process
10.1.4. >150-200: can't know everything going on, bureaucracy has to exist
10.1.5. Startups: "You're building an F1 car and you're driving it at the same time, and everyone's arguing whether you should have a steering wheel"
10.2. Companies differ by industry
10.2.1. Insurance
10.2.2. Finance
10.2.2.1. Performance matters!
10.3. Advice for data scientists
10.3.1. Haven't applied in an arXiv paper in 12 years: still using regression because it works
10.3.2. "You're going to be cleaning up your data"
10.3.3. know your data: it'll take a long time: six months to a year or more
10.3.3.1. if you don't know your data, you'll make a bizarre statement
10.3.4. to really know your data, you need to make friends with people with domain knowledge
11. Ryan Williams, senior data scientist at Starbucks
11.1. what you need to do to knock an interview
11.1.1. preparation!
11.1.1.1. A lot of people think they can just walk into an interview and their experience is naturally going to shine through
11.1.1.2. But unless you've really prepared, you can babble and talk in circles!
11.1.1.3. Read up on the typical questions you're going to face
11.2. what if you don't know the answer
11.2.1. Having a job is about resourcefulness in your ability to solve things you don't know
11.2.2. It's not about your ability to come into a room knowing everything you need to know already
12. Jarvis Miller, DS at Spotify
12.1. What surprised me about my first DS job
12.1.1. Improve as a writer
12.1.2. Explain contribution to the business without using jargon
12.1.3. How to break my work down into versions
12.2. How to deal with many suggestions for project improvements
12.2.1. Give ideas weights based on their contribution towards the goal and how long they would take
12.3. Advice to juniors
12.3.1. Don't be afraid to speak up
12.3.2. Your opinion is valued
13. Sade Snowden-Akintunde, DS at Etsy
13.1. It doesn't matter how smart you are if you cannot communicate concepts to non-technical stakeholders
13.2. Communicate early and repeat yourself
13.3. Say something from the very beginning
13.4. Soft skills are what are going to take you far in your career
14. Elizabeth Hunter, senior VP of tech strategy at T-mobile
14.1. Success in career has depended on relationships I took the time to establish
14.1.1. Provided them with a lot of context about me
14.1.2. As an introvert, I had to work at how to go about it