AI Engineering / Machine Learning

ML concepts & types · train/test split; first scikit-learn model. a baseline model. P1 start

exploratory analysis · visualising features. an EDA notebook. P1

linear regression · evaluation (MAE/RMSE). a regression model. P1

logistic regression, trees · metrics (accuracy/precision/recall/F1). a classifier. P1

encoding, scaling · feature selection. engineer features. P1

clustering (k-means) · dimensionality reduction (PCA). a clustering task. P1

cross-validation · hyperparameter tuning; over/underfitting. tune a model. P1

build · evaluate + report. predict SME churn / fraud signals. P1 published

concepts (neurons, layers, activation) · your first NN. a small NN. P2 start

tensors · a training loop. train a model in PyTorch

text processing · embeddings. an NLP task

how transformers/LLMs work · using LLM APIs. LLM APIs. call an LLM from code

LLM APIs

prompting from code · RAG intro (embeddings + vector search). a mini-RAG. P2

build · report. P2 published (DL/NLP)

app architecture · build a simple LLM app. an LLM app. P3 start

chunking & embeddings · vector DB + retrieval. a RAG system. P3

agent concepts · build an agent that performs a task. P3 published (deployed LLM app)

deploy an app/API · monitoring, cost, reliability. deploy. capstone build

S-A/B: end-to-end AI application. capstone

present (live demo + repo) · portfolio, interview prep, junior AI/ML readiness. Capstone published