Every group trip has the same failure mode: the plan lives in a chat, and a chat has no state. The decision you made on Tuesday is forty messages up, and nobody scrolls.
In TripSync a trip is a real object with members, dates, a destination and a budget. You create it, share the join code, and everyone who joins gets the same view of the same thing — not their own private reconstruction of it.
Because the crew is known up front, every later feature can assume it. An expense knows who could possibly have shared it. A booking knows how many people it is for. The packing list knows who has ticked what off. That is the whole reason the app is worth having over five better individual tools: they do not share a subject.
- Join
- A short code — no account hunting, no invite links that expire
- Scope
- Trip data is visible to members of that trip and nobody else
- Roles
- The organiser can edit trip settings; everyone can add and spend
Most AI itineraries are a list of famous places with no sense of geography or time. Ours starts from the constraints that ruin real days: how long things take, and how far apart they are.
You give it the destination, the dates, the size of the group and the budget. It drafts each day as morning, afternoon and evening, with the travel between stops accounted for, so day two does not quietly require you to be in two towns at once.
Then you edit it, because you will. Reorder a stop, delete an afternoon, or regenerate a single block without losing the rest of the week. The plan is a draft that the crew owns, not an output you have to accept whole.
Travel and stays live next to it: flights, trains and buses compared for your route, and curated places to stay and eat for the destination. Pick one as a group and it locks into the plan.
- Granularity
- Per-day, split into morning / afternoon / evening blocks
- Editing
- Reorder, remove, or regenerate one block at a time
- Attached
- Travel options and stays for the same destination, in the same trip
This is the part that ends friendships, so it is the part we built most carefully. Everything with a price attached — a booking made in the app, or a cash expense typed in at a beach shack — lands in one ledger.
Each entry records two things that a spreadsheet almost never does properly: who actually paid, and who the cost was shared with. Those are different questions, and conflating them is why the spreadsheet always ends up wrong. A dinner four people ate but one person paid for is not a quarter of a payment each; it is one payment and four shares.
From those two facts TripSync computes a net balance per person — what you paid, minus what you owe. Positive means the group owes you. Negative means you owe the group. It recomputes on every change, so the number is never stale and nobody has to “do the sheet” at the end.
The last step is the one people get wrong by hand: turning a set of balances into actual payments. TripSync reduces the debt graph to a minimal set of transfers, so you settle in three payments rather than six or twelve. The worked example is in the next section — it is not a claim, it is arithmetic.
- Splits
- Equal, or exact per-person amounts — to the paisa
- Sources
- In-app bookings and manual expenses land in the same ledger
- Settle-up
- Record-only: you mark UPI / card / wallet as paid. We move no money
“Where do we eat tonight, we are near Anjuna and two of us don’t eat meat.” That is the real query, and it is not one a search box handles well.
The local guide answers in the context of the trip you are on: where you are, when you are there, and what the crew has already picked. Ask for cafés, hidden gems, the best time to see something, or how to get across town and roughly what it should cost.
Answers are grounded in the curated place data behind the destination rather than freely invented, which is the difference between a recommendation and a hallucination with a confident tone.
The AI layer is versioned separately from the app, so the guide and the itinerary builder move to newer models without anyone shipping an app update.
- Context
- Knows the destination, the dates, and the trip’s saved places
- Grounding
- Answers drawn from curated place data, not free invention
- Upgrades
- Model changes ship server-side — no app update required
Four people, one speaker, zero adapters. Packing is a coordination problem dressed up as a personal one.
The packing list is shared and categorised, and each item carries a packed state. So the group can see that the tripod is claimed and the first-aid kit is not, before anyone is at the airport.
It is a small feature. It is in here because the entire premise of the app is that the trip is one shared object — and if that is true, the list of things the trip requires belongs to the trip, not to four separate notes apps.
- Structure
- Items grouped by category, each with a packed / unpacked state
- Shared
- One list per trip — everyone sees the same ticks
Trips do not usually end. They just stop, with an unresolved balance and photos scattered across four phones.
The recap closes it. Days travelled, places visited, total spent, who paid the most, the final settlement, and the shared album — assembled into one card the group can keep.
It also does something quietly useful: it makes the money conversation a summary rather than an argument. Everyone sees the same final numbers at the same time, and the last three payments are already listed.
- Contains
- Route, days, places, total spend, top payer, final settlement
- Album
- Everyone’s photos from the trip, in one place
- Shareable
- One card, sent to the group