102 videos · ~28 hours · free
Video series
Everything on the channel, grouped so you can follow a series from episode one instead of hunting through an upload feed. Every title opens on YouTube.
The written half of the same material lives in the free CS course and the system design questions.
Subscribe on YouTubeSeries — watch in order
AI Engineering from Scratch
14 episodes · 1h 28m
Build AI applications from scratch in Python, one hands-on episode at a time.
- Ep 1Your First LLM Call in Python — for FREE (no OpenAI bill)4m
- Ep 2Stop Hard-Coding LLMs: 100+ Models, One Line of Python (LiteLLM)6m
- Ep 3Build Your Own LLM Gateway in Python — One Endpoint, Every Model (FastAPI)6m
- Ep 4Stream LLM Responses Like ChatGPT — SSE in ~10 Lines (Python + FastAPI)5m
- Ep 5Same Answer, 20× the Cost — Why Your LLM Bill Explodes (Tokens & Pricing)6m
- Ep 6Reliable JSON from Any LLM — Stop Your Parser From Crashing (Python + Pydantic)6m
- Ep 7Function Calling Explained — Let an LLM Run Your Python (Tools, the Agent Loop)10m
- Ep 8Multi-User LLM Gateway — API Keys, Rate Limits & Budgets (Python + FastAPI)6m
- Ep 9Text Into Vectors — Embeddings & Semantic Search from Scratch (Python)9m
- Ep 10RAG From Scratch (Part 1) — Retrieval: Chunk, Embed & Search Your Docs (Python)6m
- Ep 11RAG From Scratch (Part 2) — Grounded, Cited Answers From Your Docs (Python)6m
- Ep 12Fix Your RAG Retrieval — Chunk Overlap + MMR Reranking From Scratch (Python)6m
- Ep 13Is Your RAG Any Good? Evaluate It — Recall@k + LLM-as-Judge (Python)7m
- Ep 14We Open-Sourced the Gateway — Run the Real Thing + Live Dashboard (Series Finale)5m
AI Agents from Scratch
12 episodes · 1h 51m
Build an agent from the loop up — tools, memory, routing — with no framework.
- TrailerAI Agents from Scratch · Build Agents in Python (No Framework)1m
- Ep 1What IS an Agent? Build the Minimal Agent Loop From Scratch (Python)8m
- Ep 3Build a Web-Search Agent From Scratch — Let Your AI Research the Live Web (Python)11m
- Ep 4Give Your AI Agent a Toolbox — Tool Routing From Scratch (calculator + web + RAG)9m
- Ep 5Give Your AI Agent a Memory — Short-Term + Long-Term From Scratch (Python)7m
- Ep 6Planning Agents From Scratch — Plan-and-Execute vs ReAct (Python)8m
- Ep 7Markdown Memory From Scratch — Files + an Index Instead of Vectors (Python)8m
- Ep 8Multi-Agent From Scratch — One Orchestrator, Three Workers (Python)9m
- Ep 9Reflection — Agents That Critique Their Own Work (Python)9m
- Ep 10MCP Just Deleted Its Handshake — Build an Agent on the Stateless Protocol (FastMCP, Python)13m
- Ep 11Guardrails, Cost Caps & Evals — Never Restrict an AI Agent by Asking It Nicely (Python)14m
- Ep 12The Finale — Ship a Real AI Agent, Then Watch a Framework Do It in 30 Lines (Python)14m
System Design Interviews
2 episodes · 19 min
Full mock system design interviews. Two voices, one whiteboard, one question per episode.
- Ep 1Design a URL Shortener — A Full Mock System Design Interview10m
- Ep 2Design Search Autocomplete — A Full Mock System Design Interview9m
More videos
CS and System Design
28 videos · 5h 3m →The concepts behind the code — one system design idea per episode, explained in detail.
- How Safe Is a 60-Minute Signed URL? — Explained in Detail14m
- Architecture Patterns: The Six Shapes a System Can Take — Explained in Detail16m
- SLIs, SLOs and Error Budgets — Explained in Detail13m
- Chaos Engineering vs Load Testing — Explained in Detail11m
- Big-O Without the Maths — Explained in Detail17m
- Hot Keys in a Cache — Explained in Detail14m
- API Authentication End to End — Explained in Detail17m
- Background Jobs Without a Broker — Postgres Job Queues Explained in Detail15m
- Heaps and Priority Queues — Explained in Detail15m
- Arrays vs Linked Lists — Explained in Detail14m
- Active-Active vs Active-Passive — Explained in Detail11m
- Hash Tables — Explained in Detail13m
- Write-Through vs Write-Back Caching — Explained in Detail11m
- Circuit Breaker — Explained in Detail9m
- Monitoring vs Observability — Explained in Detail10m
- gRPC vs REST — Explained in Detail7m
- Stateless vs Stateful — Explained in Detail6m
- Forward Proxy vs Reverse Proxy — Explained in Detail11m
- Docker vs VM — What Actually Happens Inside the Machine (Kernel, Namespaces, cgroups)7m
- 3 Articles Live For 3 Weeks. Search Never Saw Them. (The Dual-Write Problem)13m
- CORS Explained: The Attack It Stops, and the One Line That Fixes It7m
- Distributed Tracing, from Scratch: Which Service Ate the 3 Seconds?8m
- SQLite FTS5: How Full-Text Search Actually Works (Inverted Index + BM25)10m
- Cookies vs Tokens Explained — Where Your Auth Lives (CSRF vs XSS)6m
- Next-Token Prediction: How an AI Actually Writes Text (Not Magic — Just Probability)7m
- 5,000,000 Rows. 3 Steps. Here's What's Actually Inside a Database Index.13m
- 500 LLM Calls Took 17 Minutes. Here Are the 3 Lines That Fixed It.6m
- Authentication vs Authorization: AuthN vs AuthZ (and why 401 ≠ 403)2m
AI Engineering Deep Dives
26 videos · 3h 48m →Short, punchy deep dives on AI engineering — RAG, transformers, LLM internals.
- Webhooks — Explained in Detail18m
- The Ring Buffer — Explained in Detail15m
- 25 Years of Object Detection: Sliding Window → CNN → YOLO → DETR17m
- Why Your RAG Demo Breaks in Production — Evals, Reranking, MMR vs VRSD8m
- Stop Following Prompt Threads — Anthropic's Real 6 Themes Explained9m
- what is an Ai agent #llm #machinelearning #aiagents#langgraph1m
- LangChain in 15 Lines (And the 30 Lines It's Hiding From You)9m
- LangChain Explained — Your First AI Agent in 7 min8m
- Production LLM Guardrails9m
- MLflow Is the Wrong Default — A Model Registry Fits in 200 Lines of FastAPI8m
- Your Prompt Change Just Broke Production — Here's How to Catch It8m
- MCP's Hidden Cost: 65 Tools Burn 11,700 Tokens Before the User Says Hi7m
- The MCP Hype Is Wrong — Here's How AI Agents Actually Read Your Data7m
- Your RAG Is Broken One Step Earlier Than You Think8m
- Every RAG Demo Is Lying to You — Hybrid Search Explained7m
- OpenClaw! What Does It Actually Fix?9m
- LLMs Don't Remember Anything - AI Memory Deep Dive8m
- PostGIS vs DuckDB: Which Database for Spatial Data?7m
- Choosing Your AI Model7m
- Vision Language Models: How AI Learns to See & Talk (part 3 of 3)6m
- Vision Transformers — How Transformers Learned to See (Part 2 of 3)8m
- How Transformers Actually Work - Deep Dive into AI's Most Important Architecture (part 1 of 3)7m
- RAG vs Knowledge Graphs: Which Makes AI Smarter? | AI Engineering Deep Dive6m
- AI Agent Deep Search Deep Dive - How to Build a Research Agent From Scratch | Tech Podcast7m
- Building Trustworthy AI Agents: ReAct Framework Deep Dive Podcast6m
- How LLM Function Calling Really Works - Technical Deep Dive Podcast18m
Computer Vision
20 videos · 15h 8m →Classic computer vision, worked end to end: detection, calibration, geometry.
- 25 Years of Object Detection: Sliding Window → CNN → YOLO → DETR17m
- Kubernetes Beginner To Expert Level In One Video | Ultimate Kubernetes Guide | #kubernetes #k8s420m
- Kubernetes Beginner To Expert Level In One Video | Ultimate Kubernetes Guide | #kubernetes #k8s420m
- SFM from Scratch1m
- Image Stitching1m
- Image_Segmentation & Visual effects1m
- Traffic Sign Detection1m
- FaceLandmarkDetection1m
- FaceLandmarkDetection1m
- Single_Shot_Detector for object detection1m
- Yolo_v3 object detector1m
- Scanner1m
- Structure from motion1m
- Structure from motion1m
- Camera Calibration2m
- Camera Calibration2m
- Object size estimation1m
- Lane detection Hough transform1m
- Face detection1m
- C++ Development with Visual Studio Code (C++ extension, CMake Tools extension)33m