Skip to main content
Agentic AI Hub
Get the app

Learning Paths, Courses & Books

Table of Contents
Link copied!

Organized beginner → advanced. Pick a lane and go deep. Breadth comes with time.

Roadmaps (start here to orient)

  • roadmap.sh (AI Engineer): roadmap.sh/ai-engineer. Visual, community-maintained path from fundamentals to building AI products. A good map of the territory. 🔄
  • roadmap.sh (AI/Data Scientist): roadmap.sh/ai-data-scientist. For the modeling and research side rather than the app side.
  • roadmap.sh (AI Agents): roadmap.sh/ai-agents. Design, build, and ship agents; the newest and most on-trend of the three roadmaps. 🔄

Beginner (build intuition + first models)

  • Andrew Ng, Machine Learning Specialization: coursera.org/specializations/machine-learning-introduction. The classic on-ramp to ML fundamentals.
  • DeepLearning.AI Short Courses: deeplearning.ai/short-courses. Free 1-to-2-hour project courses on RAG, agents, and fine-tuning, built with the actual tool vendors. High ROI for practitioners. 🔄
  • 3Blue1Brown, Neural Networks series: youtube.com/@3blue1brown. Watch before or alongside any course. The visual intuition sticks.
  • Book: AI Engineering, Chip Huyen (2025): oreilly.com. Intro to building applications on foundation models. A starting point for the app-builder path.
  • Book: Hands-On Machine Learning with Scikit-Learn, Keras & TensorFlow, Aurélien Géron (3rd ed., 2022): oreilly.com. The best-selling practical ML reference; code-first and thorough.

Intermediate (build real things, understand internals)

  • Andrej Karpathy, Neural Networks: Zero to Hero: karpathy.ai/zero-to-hero.html. Build a GPT from scratch, line by line. A free deep-learning course.
  • fast.ai, Practical Deep Learning for Coders: course.fast.ai. Top-down and code-first. Get models working, then learn why.
  • Hugging Face Courses: huggingface.co/learn. Free NLP, LLM, agents, and RL courses using the HF stack; very hands-on. 🔄
  • *Book: Build a Large Language Model (From Scratch), Sebastian Raschka* (2024): manning.com. Implement a full LLM yourself. The paper-to-code bridge.
  • Book: Hands-On Large Language Models, Jay Alammar & Maarten Grootendorst (2024): oreilly.com. Visual, practical guide to using and understanding LLMs.
  • Book: Natural Language Processing with Transformers, Tunstall, von Werra & Wolf (rev. ed., 2022): oreilly.com. The canonical transformers/Hugging Face book, by HF's own authors. Pairs with the HF courses above.
  • Course: Generative AI with Large Language Models (DeepLearning.AI + AWS): coursera.org/learn/generative-ai-with-llms. The de facto standard applied-LLM course: prompting, fine-tuning, and RLHF end to end.

Advanced (research depth + production scale)

  • Stanford CS224N (NLP with Deep Learning): web.stanford.edu/class/cs224n. Lectures on YouTube; the academic backbone for NLP and LLMs.
  • Stanford CS336 (Language Modeling from Scratch): stanford-cs336.github.io. Build a full LLM training stack. About as close to how frontier labs work as a course gets. 🔄
  • Book: Designing Machine Learning Systems, Chip Huyen (2022): oreilly.com. A production MLOps reference for shipping ML at scale.
  • Book: Deep Learning, Goodfellow, Bengio, Courville (2016): deeplearningbook.org. The foundational theory text, free online. Dense but canonical.
  • Read the papers: Key Papers & Reading List is itself the advanced curriculum. Reproduce one from scratch. Nothing teaches faster.

A pragmatic path for a working developer (2026): roadmap.sh/ai-engineer to orient → Karpathy Zero-to-Hero for internals → a couple of DeepLearning.AI short courses on RAG + agents → AI Engineering (Huyen) → then ship something real and let People to Follow (X/Twitter)/Communities keep you current. The half-life of specific tools is short. The fundamentals and the habit of following the frontier are what compound.

Link copied!