Ace the Data Science

Interview
I. Four Resume Principles to Live By for Data Scientists

1. Sole purpose of your resume is to land an interview

○ Keep it short
○ Highlight a few of your best achievements
○ Omit non-relevant jobs
○ Taylor to specific job and company
○ List visa holding/eligibility at the top
2. Build your resume to impress the non-technical recruiter
○ Spell out technical acronyms
○ Tell BIG numbers
3. Only include things that make you look good
○ Include GPA/ranking only if beneficial
○ Avoid neutral information (summary at the top) or irrelevant details
○ Eliminate the technical skills section, list softwares in past experience section
I. Four Resume Principles to Live By for Data Scientists

4. Break formatting and convention to your favor

○ Bold words, frases or metrics besides section titles
○ List best things at top of resume
○ Items in a section don’t have to be chronological, rank from most to less impressive
5. Miscellaneous resume hacks
○ Make name and (professional) email prominent
○ Never include mailing address or phone number (companies biased toward local candidates)
○ Include GitHub
○ Hyperlink when possible
○ Use good PDF name, listing company name (e.g., “benjamin_leiva_meta_resume”)
II. How to Make Kick-Ass Portfolio Projects
1. Majority of recruiters just read the project description for 10 seconds
2. Optimize data science portfolio to impress the decision maker:
i. Idea that makes interesting story: buildup (data exp.), conflict (hyp), resolution (result).
ii. Cool visualizations (image, GIF, interactive)
iii. Make your project about your passion (it’s contagious!)
iv. Work with interesting datasets (no common ones, scrape it yourself, etc.)
v. Done > Perfect; turn the data science analysis into a product (post code on GitHub, web app)
vi. Demonstrate business value, point to relevant metrics ($$, visits, downloads, time saved,
etc.)
3. Double down on one portfolio
i. Choose one project that encapsulates all behavioral questions:
○ “Show, not tell”
○ “Tell me about a time…”
III. Cold Email Your Way to Your Dream Job in Data
1. Who to contact, by company size:
i. Small (~50): CEO/CTO
ii. Mid (~50-250): Technical/Regular recruiter
iii. Large (+1000): University/Campus recruiter
iv. FAANG: Data/ML recruiter
v. Bonus: alumni from your school that works there
2. Tips for cold email:
i. Keep it short (10 seconds to read, 50-125 words, avoid “hope you’re doing well” messages)
ii. Mention 1-2 accomplishments (highly relevant accomp. or internship exp.)
iii. Add urgency and establish timeline (if gotten a competing job offer)
iv. Relate personally to the recruiter/company
v. Have a specific ask (e.g., set an interview or technical test)
vi. Have strong email subject line (e.g., Former X Intern with Upcoming Deadline interested in Y)
vii. Follow up three times (day 0, day 3-4, day 8-9)
viii. Send email at the right time (Tuesday-Thursday, 11am or 2pm)
IV.Ace the Behavioral Interview
1. When do behavioral interviews happen? All the time, in every interview!
2. Three things behavioral interviews test for:
i. Soft skills: how well do you communicate, how do you work with others
ii. Position fit: how interested are you in the job/team? What motivates you about the position?
iii. Culture fit: How well do you fit the team and company’s culture?
3. “Tell me about yourself”
i. Limit answer to 1-2 minutes, start your story at a strategically relevant point (e.g., graduation)
ii. Relate your story to the position and company
iii. Mention a big accomplishment or two
iv. Rehearse!
v. Answer should include: who you are, how you came to be where you are today (sprinkle in
achievements here), what you’re looking for now (i.e., this role + company)
4. Why did you choose Data Science?
i. If path isn’t straightforward, show you’re a go-getter
ii. Talk about how skills from prior jobs/industry naturally lead you into data science
IV.Ace the Behavioral Interview
5. “Tell me about a time you…”
i. Dealt with a setback - how did you handle it?
ii. Had a deal with a particularly difficult co-worker - how did you manage it?
iii. Made a decision that wasn’t popular - how did you go about implementing it?
iv. Accomplished something that made you proud - why was that moment meaningful to you?
v. You missed a big deadline - how did you handle it?
6. “Tell me about a time…” (for data scientists)
i. When data helped drive a business decision
ii. Where the results of your analysis were much different than what you would have expected.
Why was that? What did you do?
iii. When you had to make a decision BUT the data you needed wasn’t available
iv. You had an interesting hypothesis - how did you validate it?
v. When you disagreed with a PM or engineer.
IV.Ace the Behavioral Interview
7. Use STAR approach for answering behavioral questions:
i. Situation: Describe a specific challenge you/team/company/customers encountered
ii. Task: Describe the goal you needed to accomplish (project/task)
iii. Action: Describe your role in the project/task using first person (not what your team did)
iv. Result: Describe what happened as a result of your actions. What did you learn/accomplish?
8. Ace project walk-through questions:
i. How did you collect and clean the data? Did you run into any issues when interpreting data?
ii. How did you decide what models and techniques to use? What did you eventually try?
iii. How did you evaluate the success of your projects? Was there a baseline to compare against? What
metrics did you use to quantify the project’s impact?
iv. Did you deploy the final solution? What challenges did you face launching your work?
v. What tough technical problems did you face - and how did you overcome them?
vi. How did you work with stakeholders and teammates to ensure the project was successful? If there
were any conflicts, how did you resolve them?
vii. If you did the project again, what would you do differently?
IV.Ace the Behavioral Interview
9. “Do you have any questions for us?”
i. “I don’t really have any” is NOT the right answer
ii. Follow the “until I have the offer in hand, I need to keep showing why I’m a good fit” mindset
iii. Prepare at least 3 smart, interesting questions per interviewer. Don’t pass on this!
iv. This is the time to leverage the company and product research you did.
v. Think about internal details, design decisions, or what’s coming next for some product.
vi. Ask personal questions to the interviewer:
○ How did you come into this role or company? What’s the most interesting project you’ve worked
on? What do you think is the most exciting opportunity for the company or product? In your
opinion, what are the top three challenges facing the business? What do you think is the hardest
part of this role? How do you see the company values in action during you day-to-day work?
10. Post-interview etiquette
i. Interview is NOT over! Send a follow-up thank you note via email a few hours after your interview to
keep your name and abilities fresh in their mind.
ii. Mention a few specific things you connected with them over the interview
V. Probability
1. Conditional probability
i. Probability of an event A given that an event B has occurred.
ii. Bayes’ rule:

Where is the prior, is the likelihood, and is the posterior.

iii. If then A and B are independent (i.e., B gives no information about A)

iv. A and B can be conditionally independent under event C:

v. In an interview, if other information is available and you are asked to calculate a probability,
you should always consider using Bayes’ rule (e.g., using of phrases like “given that”)
V. Probability
2. Law of total probability
i. Assume we have several disjoint events within B having occurred, then the probability of
event A happening is:

ii. can be decomposed into a weighted sum of conditional probabilities based on each possible
scenario having occurred.

iii. When asked to assess a probability involving a “tree of outcomes” upon which the probability
depends, be sure to remember this law.
V. Probability
3. Counting
i. “How many ways can five people sit around a lunch table?”
ii. “What is the likelihood that I draw four cards of the same suit?”
iii. Remember when selections does or doesn’t matter.
iv. If the order of selection of the n items being counted k at a time matters, then use:

v. If the order of selection doesn’t matter, then use:

vi. Real life applications of both: making up passwords (order matters) or choosing restaurants
nearby on a map (order doesn’t matter).
V. Probability
4. Random Variables (RV)
i. Is a quantity with an associated probability distribution. They can be discrete or continuous.
ii. If discrete, it has a PMF and can take particular values with a particular probability. If continuous, it has a PDF
and the probability of a particular value isn’t measurable. They must add up to one:

Discrete: ; Continuous:

iii. The CDF is often used instead of PMF/PDF. It’s non-negative and monotonically increasing, and its defined
as:

iv. Whenever asked about evaluating a RV, identify both its PDF and CDF.

5. Joint, Marginal, and Conditional Probability Distributions

i. RV are often analyzed with respect to other RV, giving rise to joint PDFs/PMFs.
ii. When asked about more than one RV, make it a point to think in terms of joint distributions.
V. Probability
6. Binomial Distribution (Discrete)
i. Gives the probability of k number of successes in n independent trials, where each trial has a
probability p of success. Its PMF is:

with mean , and variance

ii. Examples: coin flips, user signups, any situation involving counting successful events of
binary outcomes
V. Probability
7. Poisson Distribution (Discrete)
i. Gives the probability of the number of events occurring within a particular fixed interval where
the known, constant rate of each event’s occurrence is . Its PMF is:

with mean , and variance

ii. Examples: number of visits to a website in a certain period of time, number of defects in a
square foot of fabric.
V. Probability
8. Uniform Distribution (Continuous)
i. Assumes a constant probability of an X falling between values on the interval a to b. Its PDF
is

with mean , and variance

ii. Examples: sampling (e.g., random number generation), hypothesis testing cases.
V. Probability
9. Exponential Distribution (Continuous)
i. Gives the probability of the interval length between events of a Poisson process having a set
rate parameter of . Its PMF is:

with mean , and variance

ii. Examples: wait times (e.g., time until customer makes a purchase or the time until a default
in credit occurs).
V. Probability
10. Normal Distribution (Continuous)
i. Distributes probability according to the well-known bell curve over a range of X’s. Given a
particular mean and variance, its PMF is:

with mean , and variance

ii. Examples: many applications involve the normal distribution, largely due to (a) its natural fit
to many real-life occurrences, and (b) the Central Limit Theorem (CLT).
V. Probability
11. Markov Chains
i. It’s a process in which there is a finite set of states, and the probability of being in a particular
state is only dependent on the previous state. In other words, given the current state, the past
and future states it will occupy are conditionally independent.
ii. The probability of transitioning from state i to state j at any given time is given by a transition
matrix, denoted by P:
VI. Statistics
1. A