ItineraryTrip.com

Our AI-driven platform help you to get itineraries for any country, tailored to your interests. Save travel tips and extend your itineraries with our AI assistant.

Comment

Inspiration

As someone who travels frequently, I aimed to minimize preparation time, focusing on efficient time distribution, selecting destinations, and discerning must-see attractions from those that can be skipped.. Sure you can find all that browsing the web and opening hundreds of pages, reviews and blogs but I'm lazy :)

I wanted to quickly access to any itinerary around the world: chose a profile a country in the world and a duration

The itinerary was just the start of the journey ( no pun intended!) . When you plan a travel you want to know more about the place. There are lot of questions you might be interested in. Those questions could be related with the country, or with the cities that made the itinerary. Did i say i was lazy? Ok yes I am, so I was aiming for a chat experience with the AI where the user does not need to type a single thing. So I handcrafted all the possible questions that i could imagine organized in topics, so everything is "click-driven".

My mind started to think on what else I would like to have. If a itinerary is made of different places and activities , would not be great to do a deep dive on each of the activities? Yup, done that

Then I thought that when I collect information during the preparation phase , most of the time, its results end up in a tab which is closed, or in the best case scenario as a simple note in some handbook ...... So then I decided to provide the ability to "save" any question or activity result as a Travel note. It could be of good use once you are on the city you are visiting, or weeks/days before

Lastly but not least important I made some focus on personalization for registered users, being able to chat in different languages, listen the chat results and of course provide the sharing social features that would make my humble website viral ;)

What it does

ItineraryTrip aims to provide a enhanced platform where travelers can access nearly everything they need in one single place.

When we plan a trip is that you need to keep track of tons of links here and there. Would it not be great to have a single place to keep them all? Our 'click-driven' user experience eliminates the need for typing, making information gathering intuitive and effortless. Itineraries, maps, questions about the country , questions about the city, detail insights on activities, is just one click away

Every piece of advice from our AI becomes a personalised Travel note, easily saved for future reference

Features

  • Plan your trip: Chose your duration, anywhere in the world, any traveler profile
  • Browse our one-day visit to cities around the world
  • Explore weekend getaways for two or three days on cities
  • Browse itineraries on most countries for one and two weeks
  • Free 5 AI credits so you can explore our services / Unlimited credits with subscription plan
  • Access to attractions
  • Share your plans in social media
  • Some affiliate links to book activities on the cities of your itinerary
  • Save itineraries & Download as PDF
  • AI Travel Assistant: Personalised recommendations
  • AI Travel Assistant: Chat in multiple languages & Audio notes
  • AI Travel Assistant: Ask anything about cities and countries
  • AI Travel Assistant: Ask about any activity part of the itinerary.
  • AI Travel Assistant: Let our agent suggest extensions on your itineraries
  • AI Travel notes: Save questions and answer for later

How I built it

  • Uses Next.js : Static site rendering and dynamic pages
  • SST.dev + OpenNext so can be deployed in AWS.
  • Next-Auth for signup and Stripe for payments
  • A single table design in DynamoDB for user, subscriptions and credits management

Challenges I ran into

  • Learn Frontend stuff , this was my first React/NextJs project
  • Generation of data - I wanted to avoid user waiting for AI responses unnecesarily , so I created three basic profiles and went over the world wide web to ask our LLM(Language Large Model) from OpenAI for specific itineraries of different duration
  • Prompt engineering - It's and art. You need time and patience till you tame the OpenAI agent to reply exactly the way you want. I did not time to play with function calling so quite happy with the results I obtained
  • Streaming on AWS lambda is still immature, so thanks i found SST to help me for that
  • Visibility: Let's admit it - I'm Mr Nobody with less than 60 follower in Twitter/X. I never tried to expand my network in the past. So what are the chances of success after launch date if you do not have lots of follower that can see and spread the voice around? So I had to think hard on how to reach people via SEO, and that is a crowded space where lots of giants are competing for the same keywords space.
  • Business model: I may have changed the model 10 times since I had the idea: ads based, affiliate linking, credits based, suscription based, etc.

Accomplishments that I'm proud of

I solved the problem I wanted to solve and given that I do not consider myself as a frontend guy the results are okish. It renders quite well even on mobiles as i followed mobile responsive design first

What I learned

  • Frontend skills
  • Prompt engineering
  • SST skills
  • A bit of SEO
  • Success is really difficult unless you have a good marketing strategy!

It was a hell of a ride!

What's next for ItineraryTrip.com

  • Experiment with different LLMs such as Claude or Grok to test accuracy of results
  • Ability to craft itineraries based on a city instead of a whole country
Share this project:

Updates