AI Data Systems

Build reproducible data, feature, and language-model preparation pipelines.

You'll build: A versioned pipeline with deterministic transforms, split checks, and a reproducible run record.

1. Data Management~1h
2. Feature Engineering & Selection~2h
3. ML Pipelines~2h
4. Data Pipelines for Pre-Training~2h

Build context and retrieval paths whose quality can be measured independently.

You'll build: A retrieval system with a documented corpus, index configuration, and retrieval metrics.

1. Embeddings & Vector Representations~1h
2. Context Engineering: Windows, Budgets, Memory, and Retrieval~2h
3. RAG (Retrieval-Augmented Generation)~2h
4. Advanced RAG (Chunking, Reranking, Hybrid Search)~2h

Connect the data path to an evaluated production-style application.

You'll build: A traceable AI data pipeline with end-to-end quality checks and application-level evaluation.

1. Evaluation & Testing LLM Applications~1h
2. Building a Production LLM Application~2h

Show how telemetry separates data, retrieval, and generation failures.

You'll build: A failure analysis that links a production symptom to the responsible data or retrieval stage.

1. LLM Observability Stack Selection~1h