Launching today
Tripmojo syncs your booking emails into one shared trip, then Mojo, your AI travel agent, helps run it. Plan in a group chat where Mojo answers, runs polls and finds places. On travel day Mojo tries to check you in, saves your boarding pass, flags delays and gate changes, and tells you when to leave. On the trip it keeps today's plan current and splits expenses. Free on iOS and Android.













Free
Launch Team / Built With

AWS Activate CreditsCredits that fuel your startup’s next milestone
Promoted





Hey Product Hunt!
Every group trip we planned ended the same way: bookings spread across everyone's inboxes, the plan buried in a WhatsApp thread, and someone at the airport asking which terminal we're in. So we built Tripmojo.
How it works
Your trip builds itself. Connect Gmail and your flights, hotels, trains and activities sync into a day-by-day trip your whole group shares.
Plan together, with Mojo in the chat. Drop a Reel or a link and tag @Mojo. It pulls out the places, runs polls so the group can decide, keeps track of what's decided, and compares and books hotels, activities, eSIMs and airport rides right from the chat.
Mojo works in the background. It asks for missing passport details before check-in opens, then tries to check you in automatically. It watches your hotel price and lets you switch in a tap when the same room gets cheaper, nudges you when an airport transfer isn't booked, and points you to lounges your cards get you into.
Travel day, handled. Your boarding pass lands in Docs, and Mojo flags delays and gate changes with a leave-by time. If the airline won't let it check you in, it reminds you in time.
On the trip. Today's plan, spots near your hotel, visa rules for your passport, and shared expenses that settle up.
It's free on iOS and Android.
We'd love your honest feedback, especially: which booking email did we read wrong, and what would you want Mojo to do for you next?
Rishabh and the Tripmojo team
So when i logged in with my mail it made my two Flixbus trips on its own, that was very interesting part for me.
But I noticed one thing. My payment receipt PDF has Booking number: #3403482174 written inside it, same as my booking. The app even picked up the ₹388 amount from it for expenses. Still, the receipt didnot get attached to that booking. It is just lying separately in “Other documents” section
So i want to know, how do you decide which document belongs to which booking?
Do you match it with the booking number, or does the AI decide on its own?
@abhijeet_rauthan Thanks Abhijeet, sharp catch.
How it works: the AI reads each email and PDF and pulls out the booking number, and then our code, not the AI, matches documents to bookings on that number. A PDF follows the email it arrived in, so when an email creates a booking, its PDFs attach to that booking.
Your receipt came in a separate email with the same booking number. We correctly recognised that email as being about the same booking, but in that case we only attached the email itself, not the PDF inside it. So the receipt ended up in Other documents. That's a bug on our side, and it happened to both your FlixBus receipts.
We're fixing it so any document carrying a booking's number attaches to that booking, and we'll move your two receipts to their trips.
(Small note: the ₹388 in expenses comes from the booking itself, not from the receipt.)
@sahilgoel Thanks for the reply now i understood how it works. I thought AI picked the wrong number but the issue was the PDF only goes with its own email, got it. And my bad on the ₹388 part i thought it came from receipt.
One more question, when you fix this will it also fix the old receipts for other users or only for the new ones? and if one receipt has two bookings in it like a round trip will it attach to both?
@abhijeet_rauthan @abhijeet No worries at all, and thanks for digging in this far. This kind of feedback really helps us.
On old receipts: yes, the fix will cover receipts already in the app too, not just new ones. Once it ships, we'll re-run it over existing bookings, so other users' old receipts get corrected as well.
On round trips: yes, if one receipt covers two bookings, like an outbound and a return flight, it'll be attached to both.
Thanks again for testing so carefully!
@abhijeet @sahilgoel Thanks for explaining and all the best for the launch :P
Congrats on the launch!
I am flattered by the UI!!!!!
So like I tested the group chat with my dad on a Sikkim trip. I asked @Mojo for a trek on day 2 and it added Yuksom to the plan straight away. Then my dad said “no need, let’s go to some monastery” and I said “fine”, but since nobody tagged Mojo, the plan still shows the trek. When he did tag it, it picked up the context really well (day 2, monastery, Yuksom) and asked before swapping.
So how did u guys decide when Mojo edits the plan directly vs asks first? Is that a fixed rule (like new item = add, change to existing = confirm) or does the model decide? And did u guys think about Mojo noticing untagged decisions like “fine” and suggesting the update, instead of the plan quietly going stale?
@ameesh_mithanthaya Glad you like the UI, and thanks for testing it with your dad!
Your read is close. When you ask Mojo for something new, it adds it right away, credited to you. When a change would replace something already in the plan, it asks first, like the swap prompt in your screenshot. Undoing someone else's pick shouldn't happen silently.
Today Mojo only steps in when tagged, so it stays out of every conversation. But you've found the cost: a "fine" right after the plan card leaves the plan quietly out of date. Having Mojo spot a likely decision and offer the update, without acting on its own, is a great idea, and we're looking at it.
@garg_rish Thanks Rish! quick one, is the ask-before-replacing part a hard check in your code or does the model decide when to ask? I’d have guessed you’d keep that outside the LLM so it can’t skip it by mistake.
@ameesh_mithanthaya Its actually not binary, it depends on the confidence score based on the task at hand, LLM confidence for that specific task, automated derived preferences for specific type of task which user might have previously stated and then LLM can decide if it should confirm from user or directly apply it.
@sahilgoel makes sense, thanks Sahil.
Congrats on the launch!
Tried out the app this morning and connected my Gmail to test the booking import. Saw you also give separate forwarding addresses (@my.tripmojo.ai and +wallet) in the profile.
Curious how you guys handle updates and duplicates on the backend. Like if an airline sends 2-3 emails for a flight reschedule, or if two friends on the same group trip both sync the same hotel booking, how do you merge that without creating duplicate cards? Are you using regex/rule-based parsers for major airlines and OTAs with an LLM fallback, or running everything through an LLM?
@anshul_rawat4
Thanks so much for trying it, and for connecting Gmail!
We run everything through an LLM, with no per-airline regex. Templates change constantly, and the long tail of small airlines, hotels and local operators is exactly where rule-based parsers break. The model fills in a strict, typed schema for each kind of booking, so it always returns what matters: airports, local times, check-in dates and confirmation codes.
For updates, the model never reads an email on its own. It sees your existing trips alongside it and decides whether the email is a new booking, a change to one you already have, or a cancellation. That's how three reschedule emails from an airline end up updating the same flight card instead of creating three. We also keep a record of every email we've processed, so a re-sync never imports anything twice.
For duplicates, the LLM suggests and plain code decides. Every booking gets a hard identity key: flight number and departure date for flights, property and check-in date for hotels, plus the confirmation code where there is one. When you sync, we check against every trip you're part of, including shared ones. So if your friend already imported the hotel, your copy merges into their card instead of creating a second one. We deliberately never merge on "same city, same dates" alone, because that's how you end up merging bookings that only look alike.
Happy to go deeper on any of it, and if you ever spot a duplicate or a missed update, send it our way. Those are the cases we learn the most from.
Congrats on the launch !
I think the trip is the most important feature, because recommendations, bookings and check-in all depend on it. If the app can’t build the trip correctly from emails, nothing else works properly.
While using this app, I sent the same flight with three different PNRs. The first two had the same timings and were grouped under one flight, which looks right. The third confirmation code, Q9ZT4X, had a different timing (3:10 PM to 6:20 PM), but it also went under the same flight card, which still shows 2:10 PM to 4:20 PM (screenshot attached).
My guess is that flights are matched on flight number, date and route, with the PNRs stored under that flight, so the time never gets compared. Is that how it works? If so, I think a different time should be treated as a separate case. It could be updated automatically when flight data confirms the change, and the user could be asked when it can’t be confirmed. Otherwise a retimed booking quietly shows the old time, and anything scheduled from that time, like auto check-in, could run at the wrong moment.
@vaishnav_khanna Thank you, and you're right on every count. The trip is the foundation, which is why this one matters.
Your guess is how it works today: flights match on flight number, date and route, so the third PNR was added to the existing card without its time being compared. That's a bug. A booking with a different time for the same flight should be flagged as a schedule change, not quietly merged.
We're fixing it the way you describe. If the times differ, we check live flight data and update the card when it confirms the change. If it can't confirm, we ask you which time is right. Thanks for testing this so carefully.
@garg_rish Thanks for the quick reply, glad it was useful. Curious whether the same check will apply to buses and trains, which have no live status feed.
@vaishnav_khanna Good question. Yes, the same rule applies: a booking with a different time is never merged silently. We do have live running data for Indian railways, so that can confirm the change for us. Buses usually have no live feed, so there we'll ask you which time is right instead of guessing. If the operator sends an amendment email, we use that as the confirmation.
I found the automatic booking import feature interesting. I'm curious about how the system handles duplicate booking confirmations or booking changes. For example, if a user uploads a flight confirmation and later forwards an updated ticket, how does the system identify the existing booking and update the shared itinerary without creating duplicates?
@narayan_kumar7 Great question. This was one of the hardest parts to get right and took many sleepless nights!
Every new email or upload is checked against the bookings already in the trip before anything is saved. We first match on the confirmation code and then a lot of custom smart logic to determine if it's the same reservation. We read the new document as an update to the existing one.
So when you forward an updated ticket, we work out what actually changed, like a new time, a swapped leg or a cancelled segment. We apply only that change to the booking everyone in the group already sees. A plain duplicate of the same confirmation is recognised and merged. There's also a final check for near-duplicates, which catches things like the same hotel booked once through a site and once in the hotel's own email.
The tricky cases are the ones we test hardest: airlines reissuing a ticket under a new PNR, splitting one booking into two, or sending a "schedule change" email with no year in the dates. Each of those is a regression test in our pipeline, so the itinerary stays one clean copy no matter how many times the email arrives.
Congrats on the launch! @garg_rish I made a 2 day Ubud plan with @Tripmojo . It picked a hotel in Canggu while all the stops were in Ubud. It also set a tour 30 minutes after hotel check-in. You said code rejects anything that doesn’t add up. Does that check cover the full day, like travel time between the hotel and the stops? Or does it just cover single items like names and dates?
@kisansiyol Great catch, and fair challenge. Honest answer: those checks mainly cover single items, like the dates, times and names we pull from your bookings. They don't yet check a whole generated day, such as whether your hotel is near your stops or whether there's enough travel time between them.
Your plan shows exactly why they should. Canggu is over an hour from Ubud, and a tour 30 minutes after check-in doesn't work. We're extending the checks to the full day: hotel distance to the stops, realistic travel time between items, and a buffer after check-in. A plan that fails those checks will get fixed before you see it.
Thanks for testing this so carefully.