Tech Stack · The Console
THE TECH STACK CONSOLE
The whole AI stack as one walkable rack — the five build layers, plus principles, models, the playbook, and the directory. Pick a chip in the needle row or walk with the arrow keys; each panel shows the section, a live count from its own data, a sample of what’s inside, and a link straight in.
TECH STACK · LAYER
Hardware
The silicon layer — accelerator chips, high-bandwidth memory, the interconnect that lashes thousands of GPUs into one machine, and the fab supply chain underneath it all.
- Nvidia H100via GPU-cloud rental
- Nvidia H200via GPU-cloud rental
- Nvidia B200 / GB200via GPU-cloud rental
- AMD MI300X / MI325Xrentable on several neoclouds
- Google TPU (v5p, v6e/Trillium)via Google Cloud TPU
- Lambda
… and 3 more in the full layer page
OPEN THE LAYER →TECH STACK · LAYER
Infrastructure
The physical plant — data centers, power and cooling, the cluster network, the clouds that rent it out, and the scheduling software that keeps expensive silicon busy.
- AWS
- Microsoft Azure
- Google Cloud (GCP)
- Oracle Cloud (OCI)
- CoreWeave
- Lambda
… and 3 more in the full layer page
OPEN THE LAYER →TECH STACK · LAYER
Foundation
The models themselves — LLMs, embeddings, pretraining and alignment, quantization, multimodal, and the open-versus-closed-weights economics.
- OpenAI — GPT family + o-series reasoning models
- Anthropic — Claude family
- Google — Gemini family
- Amazon (Nova) + Bedrock
- Microsoft Azure OpenAI
- Mistral AIAPI tier
… and 14 more in the full layer page
OPEN THE LAYER →TECH STACK · LAYER
Orchestration
The wiring around the model — RAG, agents, tools, memory, vector databases, evals, and the context engineering that turns rented intelligence into products.
- LangChain
- LlamaIndex
- Pinecone
- Weaviate
- Qdrant
- LangGraphvia LangChain
… and 6 more in the full layer page
OPEN THE LAYER →TECH STACK · LAYER
Application
What ships on top — copilots, AI search, coding assistants, creative tools, agents-as-products, and the vertical apps where the stack finally meets a user.
- Cursor
- Claude Code
- GitHub Copilot
- Windsurf
- Replit
- Claude / ChatGPT (the frontier chats)
… and 29 more in the full layer page
OPEN THE LAYER →TECH STACK · GROUNDWORK
Principles
How AI actually works, from first principles — the ideas under every layer of the stack, in plain language.
- AI Principles8 ch
- Hardware4 ch
- Infrastructure4 ch
- Foundation7 ch
- Orchestration8 ch
- Application3 ch
TECH STACK · REGISTRY
Models
The model registry — the families you actually choose between, and what each one is for.
- GPT
- o-series (reasoning)
- Claude
- Gemini
- Llama
- Mistral / Mixtral
… and 6 more in the registry
OPEN THE REGISTRY →TECH STACK · GUIDE
Playbook
How to build with it: patterns, prompt craft, integrations, and the coding tools.
- Tips, Tricks & Best Practices
- The Prompt Library
- The How-To Guides
- INTEGRATIONS
- IDEs & Platforms
TECH STACK · DIRECTORY
Resources
The directory — tools, models, and where to get them, shelved by layer.