Our shared tools — a quick guide

Hey love — here's a short walkthrough of the apps we're going to use together. The goal is to spend less time planning and more time living, and to let AI help us with the boring coordination stuff (trips, budgets, household tasks, paperwork).

Each section explains: what the app is, what we use it for, how sharing works, and what you'll need to do. There's nothing scary here — most of these you already use, or are new but Mac/iPhone-friendly.

I'll handle all the billing and any tricky setup. You won't need a credit card for anything in this doc.

Part 2 — Billing engine architecture for tree-shaped products

Second half of the session: 1:1 resource→line-item model breaking under product trees, the Charge model, four-stage rating pipeline, rate-context cascade, derived charges, Burstable Internet worked example, build-vs-buy analysis, and a one-month tactical plan.

User

So antother things... until now we have had a system that loooks like this: we process audit entries to create usage recrods which have lifecycle (ordered, created, connected, etc...). Each usage record represents a resource type. Some attrs of that resource can be used to lookup a SKU which maps to prices over time. The nice thing was we had 1:1 resource to usage record and usage record to billable thing/line item on an invoice. It was all simple. Even stuff like port groups which had children were just broken into individual ports and billed as a normal port. Links and Connections, our complex types which have A and Z sides (VXCs) even stay simple because the A and Z are references so everything just gets billed on its own. I created an engine for generic pricing modifications on these usage records. Generic matcher and generic action applied to price. Multiple types of price modifications, promo codes, reseller discounts, etc... each pricing mod type could be dynamic or "grandfathered". If it's grandfathered, the mods active at order time never expire and price never changes (we always grandfather prices). If a price mod is dynamic (like reseller discounts, if they hit new tiers their discounts go up or down on existing stuff) I used clever algorithms to split the timeframe of the input into discrete chunks with different mods applied/prices. These then layer and the rules about layering the two mod types are handled properly so it can be done in any order with as many mod types as we want to exist (admin overrides, custom plans from resellers, etc...). So at the end of the day, the input would be a usage record and some time frame and you'd get back a bunch of time segments with various mods applied for each segment resulting in different prices. Cool. Math works out, everything was great. Now the 1:1 broke. We have complex products which end up being product trees. I have modeled the trees right and orchestration algos are going well. The pricing is proving tough. Now a tree with N nodes might result in <N lines (non-billable stuff, easier) or >N lines (surcharges, prices not belonging to any particular thing). I have protocols for all the node types... whether it's a resource (tracked with a lifecycle from audit logs), Locatable (has a location), Priceable (has args which map to a SKU which prices can be fetched for), etc... The question is: how do I make this work or think of it generically? Our old tabular reports won't really capture nesting, not a buge deal. What happens when mods apply to different parts of the thing? How do you split? You could have a ton of full copies of the full tree each with different prices? And some things can now come and go. Before, once something was created, it maintained the same shape. Now, nodes can come and go. Technically a port group could have ports added and removed but because we priced them by treating everything as its own thing, but now, we can have a Link which has an Internet Z side which has one or more Network spaces. Those network spaces have various sizes and therefore are priced differently and based on how long they are alive/billable. And everything can be modified. How to I handle all of this sanely? What do people do?

Public Gist

This is a public gist on my self host gist site.

# Factorial Function example

def factorial(n: int) -> int:
   ''' Factorial, returns `n!`. Equal to `1*2*3*...n`. '''
    if n < 0: