Python

A personal knowledge base of Python for data work — courses, notebooks, and code examples, organized by domain. Each category below mixes finished/authored notebooks with reference material from books and courses; sub-folder READMEs (where they exist) go into more detail.

Self-learner? Start at CURRICULUM.md — subject map, prerequisites, and three learning paths (AI Engineering, Data Engineering, Full-Stack Data). Interview prep? Start at interviewing/guides/00-start-here.md — ten pillars (+ Pillar 0 programming), ordered so each builds on the last.

Interview prep (extended): interviewing KB — study guides, round-by-round prep, and the role × topic matrix. Hands-on AI-engineering evidence (RAG, agents, evals, MCP) lives in the working repos — see ../PORTFOLIO.md. System-design writeups of those systems and a contract-based code-review drill are compiled in the librarian wiki (system-design-*.md, code-review-drill-sanyi.md).

Python for Data Analysis (data-analytics/)

• Python Basics

  • Operations
  • Texts
  • Arrays
  • DataFrames
  • Plotting
  • Aggregation
  • Time Series

• Text Analytics with Python

  • NLP Basics
  • Processing and Understanding Text
  • Feature Engineering
  • Text Clssification
  • Text Summarization and Topic Modeling
  • Text Similarity and Clustering
  • Semantic Analysis
  • Sentiment Analysis

Python for Data Engineering (data-engineering/)

• DataTalks Data Engineering

  • Docker
  • Workflow
  • DWH
  • Batch
  • Streaming

• DataTalks MLOps

  • experiment tracking
  • orchestration
  • deployment
  • monitoring

Python for Data Science (data-science/)

• Intro to Machine Learning in Python

  • supervised learning
  • unsupervised learning
  • feature engineering
  • model evaluation
  • pipelines

• Andrew Ng's Machine Learning Course

  • linear regression
  • logistic regression
  • multi-class classification
  • neural networks
  • bias and variance
  • SVMs
  • K-means and PCA
  • Anomaly detection and recommender systems

• Andrew Ng's Deep Learning Course

  • CNNs
  • Hyperparameter tuning
  • Sequence models

• Bayes/ (now in data-analytics/) and Python for ML Models/ — pruned to just the authored/completed work; generic third-party clones with no personal modification were deleted (see data-analytics/README.md for what moved where).

Python for Generative AI (generative-ai/)

Seven pillars, ordered by dependency and temporal emergence. Pairs with ai-engineering/ — gen-AI builds things with LLMs; ai-engineering is the discipline that makes them reliable.

• 01 — LLM Fundamentals — what LLMs are, how they're trained, how to prompt them. Course material: intro-to-nlp/ (NLTK, TensorFlow, transformers) and nn-zero-to-hero/ (Karpathy: micrograd, makemore, nanogpt). TypeScript examples in typescript/ (Anthropic SDK: API call, structured output, function calling, multi-turn).

• 02 — RAG & Retrieval — the first killer app pattern. Course material: DeepLearning.AI RAG, Knowledge Graphs for RAG.

• 03 — Agentic Foundations — framework learning: AutoGen, LangGraph, AgenticAIFrameworks, context engineering, agent memory.

• 04 — Agentic Frameworks — framework reference: LangGraph and ADK notes, selection guides. Course material: AI-Agents-in-LangGraph, Long-Term-Agentic-Memory-With-LangGraph.

• 05 — RL & Alignment — reinforcement learning, RLHF, and how models are aligned post-pretraining.

• 06 — Observability — LangFuse tracing, scoring, and evaluation pipelines for LLM applications.

• 07 — Agentic Applications — specific built projects: internet-search agent, deep-research bot. Active project: 07-agentic-applications/chatbot/deep-research-bot/.

AI Engineering (ai-engineering/)

• Six foundations — prompt → context → harness → loop → graph → eval. Depth companion to the interviewing guides (guides summarize; this goes deep). Each pillar cross-links its guide, cleaned notes, and coursera code.

Readings (distributed by pillar)

• Reference papers and book chapters, no notes — pure reference, colocated with the pillar that uses them:
  • data-analytics/readings/ — statistics and data-viz references (ISLP, Statistics Done Wrong)
  • data-engineering/8-data-eng-data-mesh/ — data mesh and data-engineering handouts
  • generative-ai/01-llm-fundamentals/readings/ — prompting, RLHF, and LLM foundations
  • generative-ai/02-rag-retrieval/3-rag/ — retrieval-augmented generation papers
  • ai-engineering/05-graph/3-rag-knowledge-graphs/ — knowledge graphs for RAG
  • ai-engineering/readings/ai_engineering/ — AI design, AI engineering, and performance books
  • ai-engineering/readings/general/ — foundational ML/NLP papers (word2vec, LIME, interpretability)

Programming (programming/)

• Practice problems and cloud/infra reference — doesn't fit the data-domain categories above, kept as its own bucket
  • HackerRank/ — Python and SQL practice problems
  • Leet-Code/ — arrays/hashing, two pointers, sliding windows, linked lists, plus a competitive-programming reference book