Tools

Pages that compute an answer instead of describing one. Every number is calculated from the model's real architecture or from a price captured at a stated time — never typed in.

The Build Kit

21 small interactive explainers used across the learning path— a request's journey, who-can-see-what row security, a context-window meter, a temperature dial, a load simulator and more. Each lesson that uses one offers an embed snippet, so teachers and bloggers can drop the live widget into their own page with a credit link.

request-journey
A tap's journey: URL → DNS → request → server → API
rls-toggle
Who can see which rows, with Row Level Security on and off
table-visualizer
Tables, rows, columns and a query, live
context-meter
Watch a conversation fill the model's context window
sampling-dial
Temperature: the same prompt, predictable to wild
attention-heatmap
Which words the model looks at, per token
hallucination-spotter
Spot the package that doesn't exist
session-simulator
Two phones, one server: what a login token is
build-cost-calculator
Users × actions × AI calls → a monthly bill
recipe-runner
A program, one step at a time, with its memory boxes
deploy-timeline
Laptop → preview → production → your domain
latency-ladder
Distance is time: user city vs server city
load-simulator
Launch day: queues, errors, servers, caching
spec-builder
Turn a lazy wish into a spec the model can't misread
diff-reader
Read what the AI actually changed, without coding
find-the-secret
Find the API key sitting in the frontend
soc2-estimator
SOC 2 first-year cost and timeline, line by line
server-tco-calculator
Own an AI server or rent the GPU: 3-year cost and break-even
payment-fee-calculator
Stripe, Paddle, Lemon Squeezy and more: the fee on one sale, and who files the tax
crm-cost-calculator
Eight CRMs at your seat count: total cost over 1 to 3 years, and which plan has SAML SSO
private-llm-cost-calculator
What a private, EU-resident LLM costs a month: API tiers vs your own GPU