You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
[Due for payment 2026-09-04] Display New Concierge AI Feature Modal for New Expensify users at launch #90820
Expensify is working to reestablish itself as an innovative, AI-forward tool. Getting users to recognize our innovation is key to the word-of-mouth strategy.
Problem
When new Expensify users haven't yet discovered our AI features, they may encounter competitor marketing and switch tools under the false belief that Expensify lacks those capabilities.
Solution
Create a swipe-through modal as part of the launch that showcases our AI features to new users.
Proposed Modal Screens
Concierge brings you actionable Spend Analysis — Showcase the spend analysis report. Show the #admins room for admins and ConciergeDM for employees.
An Expense Assistant to do your expenses — Highlight how users can create expenses/reports.
(Beta) Build your own agents — Show the early V1 custom agents use case and prompt users to try it.
Behavior
Dismissible for all users.
Displayed the first time existing newDot users log into (so show up not sign in) Expensify once turned live.
For brand new users do not display it after the onboarding modals. Displayed it the next time they log into Expensify.
Shown to Classic (not nudged) users if they come to New Dot, but NOT shown in Classic.
$250 to @gijoe0295 (Contributor+), per Matt's own ruling on 2026-09-08. The rules give a different number in both directions, which is worth seeing before this is paid.
#94561 paid 300% for 53 files across 3 PRs. #97231 paid the 25% floor on multiple regressions
Needs a ruling
Already ruled. Confirm it still stands given the below
Matt already ruled, and the increase request was from the wrong person
@ShridharGoel asked on 2026-09-08: "Proposed amount: $750... scope expanded... to also cover the refactoring."
Matt answered the same day: "Recommend $250, standard. Deny the $750 as submitted."
Matt also disposed of the evidence behind the request: "The 7 blockers he listed are all from PR #92007... #92007 was not fixed, it was reverted."
Worth noting separately: @ShridharGoel is a Backend Contributor and takes no per-issue payment at all. Melvin's summary says so. So the $750 request was from someone not in line for a payment on this issue. @gijoe0295 is the payee, and he has not asked for anything.
Where the rules disagree with $250, in both directions
I am not proposing either of these. Recording them because the rubric produces them and a payment nobody can reconstruct is the thing this process exists to stop.
Upward, 300%.#94012 is 41 files and +1482 lines; #97222 is 40 files and +1445 lines. Together 81 files and +2927 lines across 2 PRs. Size clears both thresholds and is past double the 25-file trigger, which adds a step. Size plus multi-PR gives 200%, plus the escalation gives 300%, ie. $750.
That is the same number @ShridharGoel asked for, reached by a different route and for a different person.
Downward, 25%. Three blockers came from #94012: #95853, #95856, #95869. #95856 is rebutted, @gijoe0295: "I suggest we ignore and close that bug", @tgolen agreeing "resizing glitches aren't worth the time". The other two carry no rebuttal, so on the rules they count, and two regressions pay the 25% floor, ie. $75.
Both cannot be right and $250 is neither. If $250 is the answer, the cleanest record is a line saying the increase is denied and the two open blockers are not held against the reviewer.
@mallenexpensify I think we should not increase the price here due to the number of regressions. I appreciate more work had to be done than initially anticipated though.
Activity
danielrvidal commentedon May 15, 2026
Going to work with @ShridharGoel and @gijoe0295 on this one! I'll update and assign once we have it fully fleshed out.
danielrvidal commentedon May 18, 2026
Working through the design mocks still.
danielrvidal commentedon May 21, 2026
Working on the mocks still.
danielrvidal commentedon May 22, 2026
I put this in the room. cc @ShridharGoel @gijoe0295. Hoping to have this fully scoped by EOW next week.
danielrvidal commentedon May 22, 2026
Gave an update in the room.
melvin-bot commentedon May 25, 2026
@danielrvidal Uh oh! This issue is overdue by 2 days. Don't forget to update your issues!
danielrvidal commentedon May 25, 2026
Moving this to coding soon!
142 remaining items
ShridharGoel commentedon Sep 8, 2026
It was decided that we will do nothing for #95856 here
melvin-bot commentedon Sep 14, 2026
@ShridharGoel Huh... This is 4 days overdue. Who can take care of this?
mallenexpensify commentedon Sep 14, 2026
Claude posting on behalf of Matt:
$250 to @gijoe0295 (Contributor+), per Matt's own ruling on 2026-09-08. The rules give a different number in both directions, which is worth seeing before this is paid.
Matt already ruled, and the increase request was from the wrong person
@ShridharGoel asked on 2026-09-08: "Proposed amount: $750... scope expanded... to also cover the refactoring."
Matt answered the same day: "Recommend $250, standard. Deny the $750 as submitted."
Matt also disposed of the evidence behind the request: "The 7 blockers he listed are all from PR #92007... #92007 was not fixed, it was reverted."
Worth noting separately: @ShridharGoel is a Backend Contributor and takes no per-issue payment at all. Melvin's summary says so. So the $750 request was from someone not in line for a payment on this issue. @gijoe0295 is the payee, and he has not asked for anything.
Where the rules disagree with $250, in both directions
I am not proposing either of these. Recording them because the rubric produces them and a payment nobody can reconstruct is the thing this process exists to stop.
Upward, 300%. #94012 is 41 files and +1482 lines; #97222 is 40 files and +1445 lines. Together 81 files and +2927 lines across 2 PRs. Size clears both thresholds and is past double the 25-file trigger, which adds a step. Size plus multi-PR gives 200%, plus the escalation gives 300%, ie. $750.
That is the same number @ShridharGoel asked for, reached by a different route and for a different person.
Downward, 25%. Three blockers came from #94012: #95853, #95856, #95869. #95856 is rebutted, @gijoe0295: "I suggest we ignore and close that bug", @tgolen agreeing "resizing glitches aren't worth the time". The other two carry no rebuttal, so on the rules they count, and two regressions pay the 25% floor, ie. $75.
Both cannot be right and $250 is neither. If $250 is the answer, the cleanest record is a line saying the increase is denied and the two open blockers are not held against the reviewer.
PR facts
To pay per Matt's ruling:
Claude posting on behalf of Matt.
PAYRUBRICTEST v2mallenexpensify commentedon Sep 15, 2026
@mjasikowski is out sick, 👀 when you're back plz.
ShridharGoel commentedon Sep 15, 2026
Note that I'll also be due payment here, this did not involve BE hours.
melvin-bot commentedon Sep 16, 2026
@mjasikowski 6 days overdue. This is scarier than being forced to listen to Vogon poetry!
mjasikowski commentedon Sep 17, 2026
@mallenexpensify I think we should not increase the price here due to the number of regressions. I appreciate more work had to be done than initially anticipated though.
mallenexpensify commentedon Sep 17, 2026
Payment Summary
Contributor: @gijoe0295 due $250 via NewDot
Contributor+: @ShridharGoel due $250 via NewDot
Thx
JmillsExpensify commentedon Sep 18, 2026
$250 approved for @gijoe0295
garrettmknight commentedon Sep 18, 2026
$250 approved for @ShridharGoel