Data Science & AI (DSP Track)
Go deep into ML, NLP, and deployment — with a portfolio that opens doors.
A 3-month, graduates-and-up track that takes you from statistics to end-to-end ML systems. You build supervised and unsupervised models, deep learning models with PyTorch, NLP pipelines with transformers, and a deployed capstone with a FastAPI service behind a Streamlit app. The program is intense — designed for graduates ready to commit to a serious career in data science and AI.
Who this is for
- Graduates (B.E/B.Tech, B.Sc, B.Stat, M.Sc, MBA) with a quantitative bent.
- Working analysts or engineers who want to move into data science and ML.
- Career switchers with a strong maths/coding base who want a deep, hireable ML portfolio.
Prerequisites
- Solid Python — functions, classes, pandas, NumPy basics.
- Undergrad-level maths — calculus, linear algebra, probability.
- Some exposure to SQL and basic stats (we offer a 1-week primer if needed).
- 8+ hours/week of focused study time outside live classes.
Primary career path
Data Scientist · ML Engineer (Jr) · AI Analyst
By the end of the program you can frame a real ML problem, build and tune models, evaluate them honestly, and ship a deployed inference service behind a small web app.
What you'll walk away with
Frame ML problems — supervised vs unsupervised, regression vs classification, baselines first.
Build supervised models with scikit-learn — linear/logistic, trees, boosting, ensembles.
Train deep learning models in PyTorch — feed-forward, CNNs, RNNs/transformers.
Run NLP pipelines with Hugging Face transformers — embeddings, classification, generation.
Deploy an inference service with FastAPI + Streamlit and version it with MLflow.
The tools you'll master
| Domain | Tools |
|---|---|
| Python Data Stack | PythonpandasNumPyscikit-learn |
| Deep Learning | PyTorchLightning (intro)Hugging Face Transformers |
| NLP | transformersdatasetstokenizersspaCy |
| Stats & Math | SciPystatsmodelssympy (intro) |
| Visualisation | MatplotlibSeabornPlotly |
| MLOps | MLflowFastAPIStreamlitDocker (intro) |
| Workflow | JupyterVS CodeGitWeights & Biases (intro) |
Real things you'll ship
End-to-End ML Capstone (Deployed)
Frame → train → evaluate → deploy a model behind a FastAPI service with a Streamlit front-end, all versioned in MLflow and containerised for repeatable deploys.
NLP Project (Transformers)
A transformer-based project — classification, named entity recognition, or summarisation — using Hugging Face, with a written technical blog post and a reproducible notebook.
What you'll earn
Greenroots Data Science & AI Certificate
Awarded on completion of the capstone, the NLP project, and a final mock DS interview panel.
Cloud ML Certification (AWS/Azure/GCP) — recommended
Curriculum aligns with the AWS ML Specialty and Azure DP-100; optional prep sprint included.
5-Phase Curriculum · 3 months
The full curriculum — every phase, every module, every project, every reading — is in the downloadable PDF. We keep it out of the web page so it stays in sync with the live cohorts.
Questions, answered
Ready to start the Data Science & AI (DSP Track)?
Download the full curriculum, then book a free career audit. We'll tell you if this is the right track for you.