Leo

customer context for agents @outlit_ai | @ycombinator 🇨🇦 🇸🇻

[Leo @LeosReal]

the inception here is crazy

[Leo @LeosReal]

how are all these company context companies proving ROI?

[Leo @LeosReal]

wtf is happening to linkedin

[Josh @joshearle]

what does "needs review" look like when 200 agent decisions flow past in an hour?

every interface paradigm shipped with a design language. the gui gave us windows and buttons. the web gave us links and pages.

there's still no equivalent primitive for supervision.

[Leo @LeosReal]

the best feeling to wake up to

[Leo @LeosReal]

@charlesmendez why would they do that much due diligence just because they wanted to not accept you? just saying no would be easier.

somethings not adding up

[Adish Jain ☕️ @_adishj]

we’re launching Motion, the frontier agent for tasteful motion design.

this launch video is made entirely with Motion. 👇🏽

QT + comment "MOTION" to get 1,000 free credits.

tag @motion_so in any X post for a surprise.

now generally available at motion [dot] so.

[Leo @LeosReal]

@eshanbetrabet scotia lounge 👀 wya

[Leo @LeosReal]

what’s stopping you from prompting like this

[Leo @LeosReal]

@vanshnyc sending the ovo goons

[Leo @LeosReal]

@tomlparser where’s that 👀

[Leo @LeosReal]

@epalacs wait until you hear my sf rent

[Leo @LeosReal]

the PMs at coinbase

[vesky @veskyxd]

you mad weird for not letting me merge

[Leo @LeosReal]

@ycombinator @manicule @greptile @skyvern @supermemory @NamanBansal0611 @shreyansj lets go manicule team

[Leo @LeosReal]

@samlambert 🙏

[Leo @LeosReal]

@DBredvick tokens spent on understanding customers > tokens spent on another poc that gets shelved

[Leo @LeosReal]

@ycombinator @avimakesrobots @ArdentAI Let’s goooo

[Leo @LeosReal]

@NamanBansal0611 lmaoo not happening 😁

[Leo @LeosReal]

kick off a couple /goal commands before bed tonight

[Leo @LeosReal]

https://t.co/lIOWTAU7GF
https://t.co/xT7KjIbouM

[Leo @LeosReal]

been turning some customer ops workflows into pi agents.

the recurring pattern is that the agent needs context before it needs autonomy.

- who is this customer?
- what changed in usage?
- what happened in support?
- what did they say in slack/email?
- what facts do we know, and where did they come from?

@outlit/pi gives pi agents tools to query that customer context.

we started with a few examples:

usage decay, friction-to-churn, activation failure, and expansion readiness.

small step, but it feels like we're getting clsoer to the right abstraction.