Learning paths
Guided paths through the best of the archive: from first step to mastery, in order.
The front page of your homelab: service dashboards with Homepage, Homarr, Heimdall and Dashy; your own push notifications with ntfy and Gotify; developer utilities with IT-Tools; and website change monitoring with changedetection.io.
Take back control of your documents, notes and tasks: private cloud, document management, wikis, collaborative notes, tasks, PDF tools and personal finance, all in containers.
The guided path to build your own infrastructure with Docker: install the engine, master networking, volumes, secrets, healthchecks and resource limits, then add a dashboard, a reverse proxy, SSO and a security layer.
Run language models on your own hardware, no API and no quotas: from Ollama to serving big models on a Mac, plus vector memory for RAG.
Know what your server is doing: metrics with Prometheus, Node Exporter and cAdvisor; dashboards with Grafana; real-time monitoring with Netdata and Glances; centralized logs with Loki and live logs with Dozzle; and a lightweight monitor with Beszel.
The data layer of your infrastructure: PostgreSQL, MariaDB and Redis; S3 object storage with MinIO; sync and backups with Syncthing, Duplicati and Seafile, plus web database managers.
Connect and protect your services: ad-blocking DNS (Pi-hole, AdGuard Home), a WireGuard VPN with wg-easy, a reverse proxy with a UI (Nginx Proxy Manager), a password manager (Vaultwarden), home automation (Home Assistant), your own Git (Gitea), a mesh network (Headscale) and a private metasearch engine (SearXNG).
What an AI agent is and how it reasons: the agentic loop and ReAct, planning and task decomposition, reflection, memory, human oversight, context engineering and multi-agent patterns.
The agents that write and edit code for you: Cline and Roo Code in VS Code, OpenAI's Codex CLI, autonomous OpenHands, Goose, Gemini CLI, Windsurf and Continue.
The pieces that connect agents to the world: building your own MCP server, the self-hosted visual platforms Langflow, Flowise and Dify, integrations with Composio, web data with Firecrawl, browsing with browser-use and the OpenRouter gateway.
The libraries to build your own agents: OpenAI Agents SDK, Pydantic AI, smolagents, Google ADK, LlamaIndex, Semantic Kernel, Mastra, Agno and Haystack.
Run tool-calling models on your own hardware: Nous Hermes 3 and 4, Qwen-Agent, function calling with Ollama, which open models support tools, and the structured-output libraries Instructor, Outlines and DSPy.
What makes an agent reliable: observability with Langfuse and Helicone, evaluation with DeepEval and promptfoo, prompt caching to cut costs, sandboxing with E2B, containerized deployment and durable execution with Temporal.
The linear algebra and calculus you need before training your first network: vectors, matrices, derivatives, the chain rule, the gradient, and the basics of exponentials, logarithms and probability.
How a network measures error: MSE, MAE and Huber for regression; binary and categorical cross-entropy for classification; entropy, KL divergence and L1/L2 regularization.
From the artificial neuron to forward propagation: weights and biases, the full family of activation functions (step, sigmoid, ReLU, GELU, Swish, ELU, SELU, Softplus, Mish) and the gradient problems.
Beyond basic gradient descent: momentum, Nesterov, AdaGrad and RMSProp, Adam and AdamW, learning rate scheduling and second-order methods.
What keeps a deep network stable: weight initialization (Xavier/Glorot and He), batch and layer normalization, and dropout.
The learning engine: gradient descent, learning rate, SGD and mini-batch, step-by-step backpropagation, the computational graph, a layer's Jacobians and the loss landscape.
