@Vince how big of a deal will it be that all of our tool calls are getting scoped to 1 child in a family

diversity is good 🌈

with regards to using these examples to build voicing/tone it's not the biggest deal

yeah

because my tool call process would be much different

like maybe not different tools but I think I'd start wide, see if anything jumps out. then probably determine the order of reports based on most interesting to least

i might even not give an update for some kids on some days

I started talking to mike and oli about this, but I think it's probably a good idea to make a "problem child" family (families). that generate real-ish data but with specific bad habits

i.e. "late night Libby" or "roblox Royce"

possible to use the data program people? @oliver

I just realized that their ACTUAL first day will be an incomplete day in all likelihood. Like they will set this up at some point during the day. Then we will start collecting data. The next morning we will have an "incomplete day" which is a good opportunity to introduce the concept of "daily observations" and maybe give them a little example if the data lends itself to it. Or even "You got all set up around 7pm last night so there wasn't much data to see after that but tomorrow morning I'll have a complete day of data and let you know!"

Not actively filtering against it and some that have made it deep in the funnel are multi-child

i can't recall

and then agent script has some of that stuff

actually not really: Parent logs in for first time: “Got you logged in ✅” “Just need to sync {child}’s device” Parent links child device via QR code: “Seeing {first name}'s device now. Just a couple more steps.” Parent finishes child device setup: “Nice work! Got {first name}’s [device] all set!” “Are you ready to move on to {next child}?” Onboarding complete: “Awesome, the family is all set up. Just text me whenever you need something and I’ll make changes from here.”

we dont have a scripted first time they text with daily review but we can make that

so not currently, no

i have not scripted anything new since we updated the waitlist onboarding

(if u wanna see old version and how we onboarded them to rules and schedules: app.notion.com/p/getclearspace/Clearspace-M…?source=copy_link)

i think personally it would be cool if the agent could be more proactive in the early phases and get a pulse check on what their appetite is for rules (vs just building another tutorial in-app)

i'm thinking through the user journey of getting their first daily report right now

@Vince would you expect the current policies to be accurately representing the state of the world given the reference date/

?

i have a case of over an hour of tik tok usage and a 20 minute budget

but no logging of rule changes in the timeline