Roadmap · Specialise
Data science (Towards ML engineering)
Opens with creditsFor students and early-career engineers who want to become data scientists and are lost in long courses with no clear order. By the end you can clean real data, test ideas with statistics and experiments, build and evaluate models, and show it all in a portfolio that answers real questions.
Curated by Tauseef Fayyaz10 stages, 71 topics~10 months at a few hours a weekBy the end, you will be able to
- Load, clean, reshape and summarise a messy real dataset in Python with pandas, or in R with the tidyverse, and keep a log of every cleaning decision
- Choose the right statistical test or confidence interval for a question, and explain what the result does and does not mean in plain words
- Design an A/B test with a sample size worked out in advance, check it is healthy, and read the result without peeking or cherry-picking segments
- Build a model with a proper split, baseline, pipeline and cross-validation, choose the metric that matches the decision, and find leakage before it fools you
- Publish two or three question-driven projects with clear write-ups and rerunnable code, and know which first roles to apply for
The path
10 stages, in the order to learn them. Open one to start.
1Python for data: NumPy and pandasSet up notebooks and a clean environment, then learn the data frame work you will do every day: loading, selecting, filtering, grouping, joining and reshaping.7 topics · ~5 weeks
- 1Notebooks and a clean environmentData scientists work in notebooks: documents where code, output, charts and notes sit together and run one cell at a time.
- 2NumPy arrays and vectorised mathsNumPy is the Python library for arrays: blocks of values of one type that you do maths on all at once.
- 3Loading and inspecting a DataFrameA pandas DataFrame is a table in memory: named columns, each one a single type, with rows that line up across them.
- 4Selecting columns and filtering rowsMost questions start by narrowing the data to the rows and columns that matter, such as paid orders from the last quarter.
- 5Grouping and summarisingGrouping splits rows by one or more columns, computes a summary for each group, and combines the results into a new table: people call it split, apply, combine.
- 6Joining tablesReal data arrives in pieces: orders in one file, customers in another, products in a third.
- 7Reshaping between long and wideWide data has one column per month or category, which is how people like to read tables.
7 in this stageOpen this stage2Statistics and probability you actually useDescribe data honestly, understand distributions and uncertainty, and learn which test answers which question, so you can tell a real pattern from noise.8 topics · ~6 weeks
- 1Describing data: centre and spreadBefore any test or model, describe each column.
- 2Distributions you will meetA distribution describes how often each value of a random quantity turns up.
- 3Conditional probability and base ratesConditional probability is the chance of something given that something else happened, written P(A|B).
- 4Samples, the CLT and standard errorYou almost never see the whole population, only a sample.
- 5Confidence intervalsA confidence interval gives a range for an unknown value, such as the true average order value or conversion rate, built from your sample.
- 6Hypothesis tests and p-valuesA hypothesis test starts by assuming there is no effect, the null hypothesis, then asks how surprising your data would be if that were true.
- 7Choosing the right testThe test follows from the question and the data type.
- 8Correlation, confounders and causationCorrelation measures how two variables move together, from minus one to one.
8 in this stageOpen this stage- SQL for data workPull the data you need from a database: aggregates, joins at the right grain, window functions and CTEs, and SQL run straight from your notebook.6 topics · ~3 weeksOpens with credits
- Data cleaning and EDATurn raw data into a table you can trust, then explore it before modelling: types, missing values, duplicates, outliers and a written first pass.7 topics · ~4 weeksOpens with credits
- Visualisation and storytellingPick the chart that answers the question, draw it in code, show uncertainty honestly, and write up a finding so someone can act on it.6 topics · ~3 weeksOpens with credits
- Machine learning with scikit-learnThe core modelling loop for tabular data: frame the problem, split, beat a baseline, build a pipeline, validate and compare. Enough to model well as a data scientist.8 topics · ~6 weeksOpens with credits
- Experiments and A/B testingDesign fair experiments, size them in advance, run them without peeking, check they are healthy and read the results honestly.8 topics · ~4 weeksOpens with credits
- Feature engineering and model evaluationBuild features that expose the signal without leaking the answer, and choose metrics and thresholds that match the decision the model supports.8 topics · ~4 weeksOpens with credits
- Working with messy real dataCollect data from APIs, open government portals and the web responsibly, document where it came from, and know the biases it carries.6 topics · ~3 weeksOpens with credits
- Portfolio, Kaggle profile and first roleTurn your work into two or three projects that show judgement, use Kaggle well, and find the route into your first data role.7 topics · ~5 weeksOpens with credits
8 more stages, 56 more topics, free with an account.
You have seen the open stages of the Data science roadmap. Every topic it covers is listed above, and an account opens the rest: what each topic means, a question to check yourself, the courses, the practice and Tauseef's advice for every stage. Early access is open: sign up and your account is active shortly after.
Everything here is free to learn from. A few things, like adding a roadmap, keeping a cheat sheet or asking Tauseef a question, use credits. You get 15 when you join, your first roadmap is free, and if you are learning seriously and run out, you can ask for more. Credits only make sure the platform goes to people who are actually learning.
Build a project
A finished project proves more than a certificate.
- Start with this oneintermediate1 semester · 1 to 2 studentsReproducing a Published Machine Learning Result with an Open Report
Validation discipline and reproducibility, the habits ML engineering roles look for
- Python
- PyTorch or scikit-learn
- Jupyter or Google Colab
- Weights and Biases or MLflow
- Git and GitHub
Related on Sefism
Picked from the other libraries by shared topics.
Best practices
See all- Data Pipeline Best PracticesHow to build data pipelines you can re-run, test and trust: idempotent tasks, orchestrator hygiene, layered transformations, data tests, freshness, ownership and monitoring.
- Data Analysis Best PracticesAn analyst's workflow in SQL, pandas and Polars: framing the question, reading data correctly, joins and grain, aggregation pitfalls, reproducible notebooks, checking results and sharing findings.
- Data Modeling Best PracticesDimensional data modeling for analytics: declaring the grain, fact and dimension tables, slowly changing dimensions, dbt marts, naming, tests, contracts and a semantic layer.
Checklists
See all- ML Model Deployment ChecklistAn ML model deployment checklist: baselines, training and serving consistency, validation, model registry, canary rollout, rollback, drift monitoring and ownership, with how to verify each.
- Data Quality ChecklistA data quality checklist for warehouse tables: ownership, freshness, volume, schema, nulls, uniqueness, referential checks, anomaly monitoring and data incidents, with how to verify each.
- Analytics Dashboard ChecklistAn analytics dashboard checklist for launch day: purpose and audience, metric definitions, data freshness, correctness, performance, access, accessibility, ownership and adoption.
Project structures
See all- Machine learning project project structureFor students and analysts whose model works in a notebook and now needs to be reproducible, reviewable, served and watched. It starts from the Cookiecutter Data Science v2 layout (immutable raw data, numbered notebooks, a real Python package) and grows into versioned data with DVC, typed parameters, data validation with pandera, experiment tracking and a model registry with MLflow, a FastAPI prediction service, drift monitoring and, for a team, scheduled retraining, model cards, fairness checks and shadow deployment.
- MLOps pipeline project structureFor engineers who can train a model and now need the training to be a pipeline anyone can rerun, review and release. A uv workspace holds the pipeline package (prepare, validate, featurize, train, evaluate, register steps run by DVC), a small shared feature package, and a separate FastAPI prediction service that loads models from the MLflow registry by alias. It grows into CI that reproduces the pipeline on pull requests, scheduled retraining, approved promotion and monitoring.
- dbt project project structureA dbt project that turns raw tables into tested, documented models, using the staging, intermediate and marts layers from dbt's own structure guide. It runs on DuckDB on your laptop, so you can learn analytics engineering for free, and grows into a governed project with checks, contracts, slim CI and a hosted docs site that moves to BigQuery or Snowflake unchanged.
Career questions
See all- What is the difference between a data analyst, a data scientist and a machine learning engineer?Analysts answer business questions with SQL, spreadsheets, dashboards and reports. Data scientists add statistics, experiments and predictive models. ML engineers build and run models inside production software. Titles vary by company, so read the work in the job advert, and pick by the kind of problem you enjoy, not the title that sounds most senior.
- Should I learn Python or R for data science?For most people aiming at industry jobs, learn Python first: it covers analysis, machine learning and production code, and most data job adverts ask for it. Choose R first if you are heading into statistics, academic research, public health or survey work, where it is strong. Either way, SQL comes too, and the concepts transfer.
- How do I move into AI engineering, data science or DevOps?Pick the field by the work you want to do every day, not by the hype. Use what you already know as a bridge: developers usually reach AI engineering or DevOps fastest, analysts reach data science fastest, and a portfolio of one or two serious projects matters more than certificates.
Cheat sheets
See allWork with me
Stuck on something specific?
Writing only gets you so far. If you want an answer to your situation rather than the general case, book a session and we will work through it together. Sessions are free for approved Sefism members, and a few slots open each week.
- 1:1 Mentorship
- Fresh Graduate Mentorship
- Career Guidance
Brand deals, promos and partnerships: tauseeffayyaz@gmail.com