Stage 1 — Computing foundations
- Python, Git, Linux basics, command line, APIs, JSON, SQL.
- Basic web concepts: HTTP, authentication, databases, deployment.
- Clean code, testing, debugging, documentation.
Stage 2 — Math and data
- Linear algebra: vectors, matrices, dot products.
- Probability and statistics: distributions, expectation, variance, confidence.
- Data analysis with pandas, NumPy, visualization, SQL.
Stage 3 — Machine learning
- Regression, classification, clustering, model validation.
- Metrics: precision, recall, F1, MAE, RMSE, ROC-AUC.
- Feature engineering, pipelines, cross-validation, model selection.
Stage 4 — Deep learning
- Neural networks, backpropagation, optimization.
- PyTorch or TensorFlow.
- CNNs, transformers, embeddings, transfer learning.
Stage 5 — LLM systems
- Prompting, structured outputs, function calling, embeddings.
- RAG architecture, vector databases, document chunking, reranking.
- Evaluation, hallucination testing, safety, cost control.
Stage 6 — Agents and production
- Tool calling, workflow orchestration, human-in-the-loop approval.
- API deployment, queues, logging, monitoring, error handling.
- MLOps, security, privacy, governance, incident response.
Portfolio projects
Build a portfolio around useful systems: invoice classifier, multilingual chatbot with RAG, dashboard forecast, AI support assistant, document extraction pipeline, and a workflow agent with approval steps.