Data Science
Interview prep for Data Scientist roles at modern AI companies — grouped by the two archetypes hiring loops actually test for: product/analytics depth, and full-stack applied modeling.
6 guides·73 chapters·~10 h of reading
No.GuideFormatLengthRead
Interview prep3
- 01Full-Stack & Applied Data ScienceFor senior/staff DS roles that ship production models end-to-end at fraud / identity-verification platforms and multimodal-sensor AI companies. 17 chapters on ML fundamentals, feature engineering, imbalanced data, prompt engineering, time-series & signals, data pipelines, production ML, MLOps, and AWS data stack.Interview prep17 ch.~2 h
- 02Product & Analytics Data ScienceFor founding-stage product DS and analytics-leadership roles at AI companies. 17 chapters on SQL, experimentation (A/B, MABs, causal inference), product metrics, predictive modeling for business, defining the data/experimentation stack, dashboards, and analytics leadership.Interview prep17 ch.~2 h
- 03Senior Data Scientist, Business Analytics — Interview PrepCompany-agnostic interview prep for the first-data-hire, Business Analytics DS role at a privacy-first AI company — the kind that runs on freemium subscriptions, a usage-based API, and a staking-for-compute token, and deliberately doesn't log user data. 11 chapters: analytics without surveillance, building the data stack from scratch, freemium + API + token metrics, privacy-safe experimentation, on-chain analytics, SQL, stakeholder communication, and the 30-60-90.Interview prep11 ch.~1 h
Deep dive2
- 04Data Science for NeocloudsTwo sections in one guide. Section A: how DS / ML practitioners pick providers, run training and fine-tuning, manage experiments, and engineer cost on neocloud infrastructure. Section B: how DS teams *inside* neoclouds build pricing models, churn analysis, capacity forecasting, marketplace ranking, and anomaly detection on platform metrics.Deep dive12 ch.~2 h
- 05Venice.ai — Company ResearchA standalone research dossier on Venice.ai, the privacy-first, uncensored AI company (Erik Voorhees, VVV / DIEM token economy, $1B Series A). How the product and its no-logging privacy architecture work, where the money comes from (subscriptions + API + staking-for-compute), and an honest market evaluation — competitors, the moat question, what works and what doesn't, bull vs bear.Deep dive4 ch.~1 h
- 06Experimentation & A/B TestingA from-zero field guide to online controlled experiments. 12 chapters: the causal logic and the counterfactual, the statistics you actually need, designing and running a test, reading the results, the pitfalls that fool smart teams, advanced designs (bandits, switchback, interleaving), quasi-experiments for when you cannot randomize, and industry playbooks across consumer, marketplaces, ranking, SaaS, fintech, mobile, and health.Deep dive12 ch.~2 h