Role roadmaps / Data Scientist Role roadmap
Decide with data: experiments, models, and uncertainty. You still need SQL. You do not need a new architecture every sprint.
Also called: Applied Scientist, Research Scientist (industry)
You ship: An experiment, a model with a holdout, and a decision memo — not a demo notebook.
3–6 months of focused practice beyond SQL
Roadmap Coding Interview Question bank Switch roleData Analyst BI Analyst Data Engineer Analytics Engineer Data Scientist ML Engineer MLOps Engineer AI Engineer Data Architect Nearby roles:Data Analyst , ML Engineer , AI Engineer
→
Skill check10 questions Already working in this role? Answer honestly and the roadmap ticks what you have done, then points you at the first gap.
Take the 2-minute skill check 1 of 10
Can you do this?
1 Yes, I have done this2 Partly3 Not yet
Back Keys 1, 2 and 3 answer while the check has focus.
Result
Ticks are only added, never removed. Partly and Not yet leave a skill unticked.
Checked skills
0% · pick any skill to see its syllabus, links and practice.
Journey Map List
− Fit + Print Clear checks
RTE
Stop 1 of 16
◀ Prev StopNext Stop ▶ Auto Cruise
Click any bus stop to drive there and inspect the complete syllabus, or scroll the highway to follow the bus.
Terminal Start · Station 1 of 16
Click to stop Start · Data Scientist An experiment, a model with a holdout, and a decision memo — not a demo notebook. Inspect station brief →
Checkpoint · Station 2 of 16
Click to stop Foundations The tools you cannot fake in week one Inspect station brief →
Bus Stop · Station 3 of 16
Click to stop SQL for features Windows, point-in-time joins, and leakage as a join bug. Open Syllabus →
Bus Stop · Station 4 of 16
Click to stop NumPy / pandas fluency Vectorize, groupby, and why a loop over 10M rows is a career event. Open Syllabus →
Bus Stop · Station 5 of 16
Click to stop Inference, not vibes Hypothesis tests, confidence intervals, and power. p-hacking is an incident. Open Syllabus →
Checkpoint · Station 6 of 16
Click to stop Daily craft What you actually do on a Tuesday Inspect station brief →
Bus Stop · Station 7 of 16
Click to stop Question-first EDA Write the decision, then the plot. Segment before you model. Open Syllabus →
Bus Stop · Station 8 of 16
Click to stop Classical ML that ships Linear models, trees, calibration. Deep learning when the table is wrong. Open Syllabus →
Bus Stop · Station 9 of 16
Click to stop Experiment design Guardrails, SRM, and CUPED as vocabulary. “We launched and revenue went up” is not science. Open Syllabus →
Checkpoint · Station 10 of 16
Click to stop How you ship Production habits, not extra courses Inspect station brief →
Bus Stop · Station 11 of 16
Click to stop Evaluation without self-deception Holdout, leakage, and a metric the business already uses. Open Syllabus →
Bus Stop · Station 12 of 16
Click to stop Handoff to engineering Features as tables, not pickle files. Online/offline parity. Open Syllabus →
Checkpoint · Station 13 of 16
Click to stop Career moves How this role talks in interviews Inspect station brief →
Bus Stop · Station 14 of 16
Click to stop DS interview SQL, a modeling case, stats, and a take-home. Talk leakage out loud. Open Syllabus →
Bus Stop · Station 15 of 16
Click to stop Next role ML engineer to serve. Analyst lead to decide without models. DE if you fell in love with pipelines. Open Syllabus →
Final Destination · Station 16 of 16
Click to stop Keep practicing SQL + Python for this role Inspect station brief →
Start · Data Scientist Foundations SQL for features Windows, point-in-time joins, and leakage as a join bug. NumPy / pandas fluency Vectorize, groupby, and why a loop over 10M rows is a career event. Inference, not vibes Hypothesis tests, confidence intervals, and power. p-hacking is an incident. Daily craft Question-first EDA Write the decision, then the plot. Segment before you model. Classical ML that ships Linear models, trees, calibration. Deep learning when the table is wrong. Experiment design Guardrails, SRM, and CUPED as vocabulary. “We launched and revenue went up” is not science. How you ship Evaluation without self-deception Holdout, leakage, and a metric the business already uses. Handoff to engineering Features as tables, not pickle files. Online/offline parity. Career moves DS interview SQL, a modeling case, stats, and a take-home. Talk leakage out loud. Next role ML engineer to serve. Analyst lead to decide without models. DE if you fell in love with pipelines. Keep practicing Phase 1: Foundations SQL for features Windows, point-in-time joins, and leakage as a join bug. NumPy / pandas fluency Vectorize, groupby, and why a loop over 10M rows is a career event. Inference, not vibes Hypothesis tests, confidence intervals, and power. p-hacking is an incident. Phase 2: Daily craft Question-first EDA Write the decision, then the plot. Segment before you model. Classical ML that ships Linear models, trees, calibration. Deep learning when the table is wrong. Experiment design Guardrails, SRM, and CUPED as vocabulary. “We launched and revenue went up” is not science. Phase 3: How you ship Evaluation without self-deception Holdout, leakage, and a metric the business already uses. Handoff to engineering Features as tables, not pickle files. Online/offline parity. Phase 4: Career moves DS interview SQL, a modeling case, stats, and a take-home. Talk leakage out loud. Next role ML engineer to serve. Analyst lead to decide without models. DE if you fell in love with pipelines.
SQL for features Windows, point-in-time joins, and leakage as a join bug.
Checked automatically: you solved the linked practice problems.
Feature SQL: windows, point-in-time joins. Leakage is a join bug.
Time to learn Until leakage looks like a join bug.
Prerequisites Windows Point-in-time joins What you need SQL + Python pads A labeled dataset with as-of timestamps Concepts to study Point-in-time (as-of) joins Entity-timestamp grain of a feature table Label leakage via future aggregates Train/serve feature parity starts in SQL Tech stacks Warehouse SQL — Features as tablesPython after the extract — Not instead of SQLHow to study Write a feature that uses only data known at as_of_ts. SQL + Python problems; windows track. What to produce A PIT join A leak you can point at in SQL Pitfalls Joining today’s label to yesterday’s feature without a time predicate Notebook joins that cannot be reproduced Interview prompts This AUC looks perfect. Where is the leak in the join? Links
Previous skill Next skill
Mark as done Search DataLane All Articles Stacks Practice Learn Pages
No results. Try analyst, numpy, snowflake, or roadmap.
↑↓ navigatealt ↑↓ jump group↵ openesc closeAliases: sf, numpy, analyst, mlops