Interview Prep Hub

Full-Stack & Applied Data Science

A study guide for senior/staff Data Scientist loops where the work is shipping production models end-to-end — calibrated against the SentiLink Staff DS, Full Stack role (fraud & identity) and the Archetype AI Data Scientist role (multimodal sensor/prompt workflows). 17 interactive HTML chapters with sticky navigation, syntax-highlighted code, copy buttons, drill modes, and dark mode.

If you only have an hour

Open 00-START-HERE first. Then drill 15-interview-questions with drill mode on. Reread 16-day-of the morning of.

Drill: Interview Questions Hide-show answers · per-Q tracker Drill: Coding Problems Python · multiple approaches Core: Production ML Code, tests, monitoring, drift
Section A · Orient
00 Start Here Master index, study schedule, the reframe that separates "ML researcher" from "full-stack DS who ships." 01 The Roles Decoded SentiLink Staff DS (Full Stack) and Archetype AI Data Scientist, side by side. What each team actually does, what the stack means, what the loop is really testing. 02 Positioning From Scratch Mindset before content — how to interview honestly when you've done some production ML but not at staff scale, and how to lean on adjacent experience without overclaiming.
Section B · Core Data Science
03 ML Fundamentals Supervised models for tabular data — logistic regression, gradient boosting, calibration, evaluation under skew, model selection at staff bar. Critical 04 Feature Engineering "Inventive feature engineering" — what SentiLink means by that. Entity-level features, leakage avoidance, drift, feature stores, and the hard part: features that encode domain insight. Critical 05 Fraud & Imbalanced Data Class imbalance, cost-sensitive learning, precision-recall tradeoffs, threshold tuning at real-world fraud rates, evaluation when labels are scarce or delayed. SentiLink-flavored. Critical 06 Prompt Engineering, Applied N-shot examples, prompt templates, lens/parameter configuration, prompt-as-config patterns, iterating prompts against an eval set. Archetype-flavored. 07 Time-Series & Signals Signal processing basics, time-series features, sensor data prep, alignment and resampling, video and multimodal data prep. The Archetype "raw data to testable results" workflow. 08 Data Pipelines, Applied Data acquisition, labeling workflows (Encord and others), iterative preprocessing cycles, feature pipelines, training data versioning, the loop from raw asset to model-ready dataset. 09 Production ML Writing production code & tests, real-time inference, monitoring, retraining cadence, A/B rollouts for models, on-call considerations. The SentiLink "end-to-end ownership" bar. Critical
Section C · Coding
10 Coding Fundamentals Python patterns full-stack DS interviewers reach for — clean classes, generators, testability, vectorization vs explicit loops, useful Big-O intuition. 11 Coding Problems Drillable problems — feature engineering puzzles, imbalanced-data evaluation, ranking and aggregation, a couple of classic algos. Multiple approaches per problem; drill mode hides solutions. Drill mode
Section D · Production / Cloud
12 AWS Data Stack EC2, S3, RDS, Redshift, plus PostgreSQL as the system of record. What the SentiLink stack implies about how their work is structured, and how to discuss it credibly. 13 MLOps Applied CI/CD for models, experiment tracking, shadow deploys, canary rollouts, drift monitoring, retraining triggers — the operational glue that separates a Jupyter result from a product.
Section E · Reference & Execution
14 Domain Context Two domains, one chapter — fraud & identity verification (SentiLink) and multimodal sensor-fusion AI (Archetype). The vocabulary and stakes you need to sound credible. 15 Practice Interview Questions ~30 Q&A across ML, features, evaluation, production, prompts, signals, and domain. Drill mode hides answers; per-question practice tracker; progress saved locally. Drill mode 16 Day-Of Tactics Structural moves, traps to watch, recovery patterns, the right questions to ask them, closing statement. Reread morning of.
Study Paths
If you have 7+ days
  • Day 1: 01, 02 (orient) → 03 (ML fundamentals)
  • Day 2: 04 (features), 05 (fraud/imbalanced — for SentiLink)
  • Day 3: 06 (prompts — for Archetype), 07 (time-series & signals)
  • Day 4: 08 (pipelines), 09 (production ML)
  • Day 5: 10, 11 (coding drills on a timer)
  • Day 6: 12 (AWS), 13 (MLOps), 14 (domain context)
  • Day 7: Drill 15. Reread 16. Sleep.
If you have 2–3 days

01, 02, 03 (ML), 04 (features), the role-relevant 05 or 06, 09 (production ML), 11 (drill 4–5 problems), 15 (drill), 16. Skim everything else.

If you have < 24 hours

01, 02, 04 (features — central to both roles), 05 or 06 (whichever applies), 09 (production ML), 15 (drill all questions), 16. Skim 03, 07, 13 headings only.