Learn
Hands-on AI/ML curriculum, one lesson a day, free.
LLMs & Transformers
-
How transformer self-attention actually works, with a runnable toy
-
Tokenization pitfalls: why your LLM miscounts characters and money
-
Context windows, KV-cache, and why long prompts get expensive
-
Fine-tuning vs LoRA vs prompting: choosing the cheapest path that works
-
Structured output and JSON mode without the model going off the rails
-
Quantization in practice: running a capable model on one consumer GPU
RAG & Agentic Development
-
Chunking strategies that actually matter for retrieval quality
-
Choosing a vector database in 2026: pgvector vs the specialists
-
Hybrid search: combining BM25 keyword and dense vectors
-
Your first tool-using agent: function calling done right
-
The Model Context Protocol (MCP): connecting agents to real tools
Computer Vision
-
Object detection today: from YOLO to open-vocabulary models
-
Image segmentation with Segment Anything and its successors
-
How diffusion models generate images, explained with a runnable toy
-
Vision-language models: teaching an LLM to see
-
OCR in 2026: document understanding beyond plain text extraction
-
Deploying vision models to the edge: quantization and latency
ML Foundations & Chatbots
-
The bias-variance tradeoff, finally explained with runnable plots
-
Data leakage: the silent killer of ML projects
-
Feature stores and reproducible training pipelines
-
Model monitoring in production: drift, decay, and alerting
-
Build your first AI chatbot: from prompt to a working assistant
-
Giving your chatbot memory and personality without prompt bloat