Pathway 6: "I'm Completely New to ML" (Career Changer)

Target audience: Career changers, bootcamp graduates, and self-taught developers entering the AI field

Goal: Build a complete mental model of the LLM stack, from basic linear algebra through production agent systems, with working code at every step.

Chapter Guide

• Start Ch 00: ML and PyTorch Foundations (your starting point) your starting point: tensors, gradients, training
• Focus Ch 01: NLP and Text Representation learn how text becomes numbers
• Focus Ch 02: Tokenization and Subword Models understand how models see words and subwords
• Focus Ch 03: Sequence Models and Attention the attention mechanism that powers everything
• Focus Ch 04: The Transformer Architecture the architecture behind every modern LLM
• Focus Ch 05: Decoding and Text Generation how models actually generate text
• Skim Ch 06: Pre-training and Scaling Laws context on how large models are trained
• Skim Ch 07: The Modern LLM Landscape survey of models you will use in practice
• Skim Ch 08: Reasoning Models and Test-Time Compute understand when reasoning models add value
• Skim Ch 09: Inference Optimization context on cost and latency tradeoffs
• Focus Ch 10: Working with LLM APIs start building with APIs right away
• Focus Ch 11: Prompt Engineering the skill you will use most often
• Skim Ch 12: Hybrid ML+LLM Architectures learn when to combine classical ML with LLMs
• Skim Ch 13: Synthetic Data Generation how to create training data with LLMs
• Skim Ch 14: Fine-Tuning Fundamentals understand how models are customized
• Skim Ch 15: PEFT (LoRA, QLoRA) efficient methods for adapting models on limited hardware
• Skip Ch 16: Knowledge Distillation and Model Merging advanced training topic; return to later
• Skim Ch 17: Alignment (RLHF, DPO) understand how models learn to follow instructions
• Skip Ch 18: Interpretability research-oriented; return to later if interested
• Focus Ch 19: Embeddings and Vector Databases store and search over your own data
• Focus Ch 20: RAG build your first knowledge-grounded app
• Focus Ch 21: Conversational AI add conversation and memory to your app
• Focus Ch 22: AI Agents build your first autonomous agent
• Skim Ch 23: Tool Use and Protocols connect agents to external tools via MCP
• Skip Ch 24: Multi-Agent Systems advanced topic; return to after gaining experience
• Skip Ch 25: Specialized Agents advanced topic; return to after gaining experience
• Skim Ch 26: Agent Safety and Production Infrastructure safety guardrails before deploying agents
• Skim Ch 27: Multimodal Models see what image and audio models can do
• Skim Ch 28: LLM Applications real-world project ideas and patterns
• Focus Ch 29: Evaluation and Experiment Design learn to measure what you build
• Skim Ch 30: Observability and Monitoring track costs and errors in your applications
• Skim Ch 31: Production Engineering understand how to deploy what you build
• Focus Ch 32: Safety, Ethics and Regulation responsible AI use is essential from day one
• Skim Ch 33: Strategy, Product and ROI understand the business side of AI products
• Optional Ch 34: Emerging Architectures survey of where the field is heading
• Optional Ch 35: AI and Society broader context on AI impact and governance