261: Succeeding in Data Science with the Trichotomy of Control

Episode Summary

Executive Summary: This episode centers on Andrei Liskov’s path from business analyst and student to data scientist at Apple, with a practical playbook for landing data science roles. He explains Stoic goal-setting, the data science interview funnel, the outsized value of referrals and personal websites, and how he taught himself data science through projects, classes, books, podcasts, and media.

Main Topics: Stoic goal-setting and the trichotomy of control: Liskov explains how to focus on controllable inputs, accept uncontrollables, and set internal goals for partially controllable outcomes like promotions or elite jobs. Data science interview funnel: He breaks down typical interview stages: coding challenge, HR screen, technical screen, take-home project, on-site, and negotiation, emphasizing that formats vary by company. Referrals and networking as hiring leverage: Liskov argues that referrals and warm introductions materially improve hiring odds, and that long-term relationship-building beats cold referral requests. Personal website and online presence: He describes building a minimalist personal site to control search results, showcase projects, and support recruiter evaluation during interviews. How he taught himself data science: He outlines a multi-channel self-learning strategy combining coursework, projects, Kaggle, DataQuest, books, podcasts, communities, and media. Specialization within data science: He recommends choosing a data science subtrack—algorithms, analytics, or inference—rather than trying to prepare for the entire broad field at once.

Key Arguments: Set goals around controllable behaviors, not outcomes you cannot fully control; this reduces frustration and improves consistency. In hiring, internal goals such as improving resume quality, building a portfolio, and securing referrals are more actionable than simply aiming for a dream company. Data science interviews are heterogeneous; success depends on researching the specific company and interviewers rather than relying on a generic prep plan. Referrals often work best when they emerge from genuine relationships and long-term mentorship, not transactional outreach. A personal website helps shape first impressions, improve discoverability, and provide evidence of work quality before interviews. Deep learning comes from project-based and multi-source exposure, not from passive consumption of theory alone. Because data science is broad, candidates should narrow their target role and deepen relevant skills instead of trying to master everything.

Data Points: Podcast episode number: 261 - This is episode 261 of the Super Data Science Podcast. Data Science Go event dates: 27th, 28th, 29th of September 2019 - Conference promoted in the intro as a career-focused data science event in San Diego. Data Science Go speakers: 40 speakers - Planned speaker lineup for the conference. Expected conference attendance: over 800 attendees - Projected audience mix of beginners, practitioners, managers, and leaders. Executive track attendees last year: business decision makers from Eli Mae, Levi Strauss, Dell, Red Bull, and more - Examples cited for Data Science Go X executive track participants. Interview article views: almost 50,000 views - His article on mastering the data science interview loop went viral within months. Interview prep volume: 20 to 30 interviews - He says he interviewed with many companies during senior year. Take-home project preparation: 15 to 20 hours in 5 or 6 days - Time spent on one take-home assignment during job search. Total prep before onsite: 40 to 50 hours in one week - Combined preparation on SQL, statistics, Python, machine learning, and take-home work. Website template cost: $5 to $10 - He says his personal website was built from a cheap HTML/CSS template. Fields in Airbnb-style taxonomy: 3 categories - Algorithms, analytics, and inference used to segment data science roles. Referral/hire pathway claim: 70% of hires via referrals (stated by host) - The host cites a referral-heavy hiring market; Liskov confirms most of his own roles came via referral or direct contact.

Pivotal Quotes: "The real world often is a much better teacher and student." — Andrei Liskov: Explaining why he prioritized experiences and work over maximizing grades in university. "I think the first thing to do is to define the role that you're looking to get and break out the skills and attributes that consist of that role." — Andrei Liskov: Advising job seekers to narrow their data science focus before preparing for interviews. "I would really suggest you identify one of these areas and go deep because I've certainly fallen in this trap when I was just starting." — Andrei Liskov: On avoiding overwhelm by specializing within the broad field of data science.

Implications: For job seekers, success in data science is less about generic studying and more about targeted preparation, networking, and visible proof of work. For the industry, hiring remains relationship-driven, so personal branding and specialization matter more than ever.