Esc
↑↓ navigate   Enter open   Esc close
🤖 Data Scientist & ML Engineer

90-Day English Deep Dive for Data Scientists & ML Engineers

A comprehensive 3-month programme that takes you from solid ML vocabulary to full professional fluency in data science English. Over 13 weeks you will master statistics and machine learning terminology; develop MLOps vocabulary and data storytelling skills; build the stakeholder communication language for presenting results to non-technical audiences; and reach advanced fluency for reading research papers, giving conference talks, and discussing AI safety.

Advanced 90 days · 13 weeks · 3 phases · 1–2 hrs/week · Full role guide →
Start Week 1 →
1
Foundations
Weeks 1–4
Stats + ML core vocabulary, Jupyter/Python terminology, daily communication
2
Intermediate
Weeks 5–8
MLOps vocabulary, data storytelling, stakeholder communication, model explanation
3
Advanced
Weeks 9–13
Research paper English, conference presentations, AI safety vocabulary, leadership

Phase 1 (Weeks 1–4): Foundations

Phase 2 (Weeks 5–8): Intermediate

Phase 3 (Weeks 9–13): Advanced

Advanced phrases you will master in 90 days

data leakage
"There's data leakage in the pipeline — the validation set contains features derived from the full dataset."
hyperparameter tuning
"We ran hyperparameter tuning with Optuna — learning rate of 3e-4 and dropout of 0.3 gave the best val loss."
ablation study
"The ablation study shows that removing the attention mechanism drops accuracy by 4.2 points."
ground truth
"The ground truth labels were annotated by three independent reviewers — inter-annotator agreement is 0.87 kappa."
distributional shift
"We're seeing distributional shift in production — the model was trained on 2022 data but real traffic looks different now."
feature engineering
"Aggressive feature engineering brought the baseline F1 from 0.72 to 0.81 before we even tried a more complex model."
cold start
"The recommendation system has a cold start problem — new users get generic results until we have enough signals."
confidence interval
"The 95% confidence interval for the improvement is 1.2 to 3.8 percentage points — statistically significant."
guardrail metric
"Conversion rate is our north star, but CTR is the guardrail metric — we won't ship anything that reduces it."
model drift
"Set up monitoring for model drift — we want alerts if prediction distribution shifts more than 10% from baseline."
Data Scientist & ML Engineer — 30 Days
The 30-day foundation path — complete this before or alongside the 90-day programme
DevOps & SRE — 90 Days
Complement your ML skills with MLOps and infrastructure vocabulary
Data Scientist & ML Engineer Guide
Comprehensive reference guide for all areas of data science and ML English

Frequently asked questions

Who is the 90-day data science English deep dive designed for?

This programme is designed for data scientists and ML engineers who work in English-speaking teams or present their work to international audiences. It covers the specialised vocabulary of statistics, machine learning, MLOps, data storytelling, and the academic English needed for reading research papers and giving conference talks.

What English level is required for the 90-day data science path?

You should be at B2 level or above. The path builds from core ML vocabulary in Phase 1 to advanced academic and presentation English in Phase 3. If you can read ML research papers in English but struggle to explain your work to stakeholders, this path is ideal.

How much time per week does the 90-day path require?

Each week has 4 resources, each taking approximately 20–30 minutes. Total weekly commitment is 1.5–2 hours. You can spread this across 4 evenings, making it sustainable alongside a full-time data science or research role.

Does this path cover data storytelling in English?

Yes. Week 6 focuses specifically on data storytelling — the language for presenting findings, structuring narratives around data, communicating uncertainty, and making technical results understandable to non-technical stakeholders. This is one of the most high-value communication skills for data scientists.

What MLOps vocabulary is covered?

Week 5 covers MLOps vocabulary: model deployment language, experiment tracking terminology, feature store vocabulary, model monitoring language, data versioning terms, and the CI/CD vocabulary applied to machine learning pipelines. This is increasingly important as ML teams adopt production engineering practices.

Does the path include research paper reading English?

Yes. Week 9 focuses on the academic English used in ML research papers: abstract structure, methodology language, limitations vocabulary, statistical significance phrases, and the specific collocations used in top ML venues like NeurIPS, ICML, and ICLR.

Is AI safety and ethics vocabulary covered?

Yes. Week 11 covers AI safety and ethics vocabulary: alignment language, fairness and bias terminology, interpretability vocabulary, responsible AI phrases, and the vocabulary used in AI governance discussions. This vocabulary is increasingly required in ML roles at companies with responsible AI programmes.

Does this path help with conference presentation skills?

Yes. Week 10 focuses on conference presentation English: structuring a technical talk, Q&A language, handling challenging questions, and the vocabulary for presenting experimental results to peer audiences. This is distinct from the stakeholder communication covered in week 7.

Is there content for senior data scientists and ML leads?

Yes. Phase 3 is particularly relevant for senior practitioners: research paper English, conference presentations, AI safety vocabulary, leadership and mentoring language, and career negotiation vocabulary — all areas where senior data scientists and ML tech leads need precise English.

What should I do after completing the 90-day data science path?

After completing this path, explore the Data Scientist & ML Engineer guide at /guides/data-scientist-ml-engineer/ for reference material, practise by writing summaries of papers you read, and consider joining English-language study groups or contributing to ML discussions on forums and GitHub.

Ready for the deep dive?

Begin Week 1 and commit to 90 days of structured English mastery.

Start Week 1 → 30-day path first All learning paths