~25 video lectures
Short, focused explainers — most under fifteen minutes — embedded directly inside each week.
ByteByteGo AI Engineer Course
Welcome to the ByteByteGo AI Engineer track. Over six focused weeks you will move from the math and mechanics behind large language models all the way to shipping an end-to-end multimodal agent. Each week pairs short video lectures with architecture diagrams you can open in Excalidraw, plus hands-on exercises to cement the ideas.
Weekly modules
Section titled “Weekly modules” 🧱 Week 1 — LLM Foundations Tokens, embeddings, attention, and the transformer block. Build the mental model you'll use for every other week.
🔍 Week 2 — Retrieval-Augmented Generation Vector databases, chunking strategies, and grounding LLMs in your own knowledge with production-grade RAG pipelines.
🤖 Week 3 — Agents Tool use, planning, and multi-step reasoning loops. Wire LLMs into real systems through function calling and orchestration.
🧠 Week 4 — Thinking & Reasoning LLMs Chain-of-thought, reflection, and the new wave of reasoning models. When to trade latency for depth — and when not to.
🎨 Week 5 — Image & Video Generation Diffusion models, control nets, and multimodal pipelines that turn prompts into pixels, frames, and full scenes.
🚀 Week 6 — Capstone Project Bring every layer together. Ship a working multimodal agent, complete with retrieval, tools, and a clean front end.
What to expect
Section titled “What to expect”10 architecture diagrams
Excalidraw source files for every major system, in both landscape and vertical layouts.
Hands-on exercises
Notebooks, prompts, and small builds that turn the lecture material into real intuition.
~5–8 hrs/week
Designed to fit alongside a full-time engineering role without crushing your weekends.
Ready to start? Open Week 1 — LLM Foundations and dive in.