Explorologie

• 

  Landing page

• 

  Event categorization features

• 

  City page

Inspiration-

I get many event emails, newsletters with calendars and event suggestions from apps. It is too difficult to keep up with, so I want to sync all these sources and better organize my time by using a physical mapping agent that embeds all of these sources into one streamlined experience. When traveling I have also wanted to simplify my itinerary planning based on what attractions and events were closer to one another to conserve time used in transit. Also, when I want to attend several events in one evening in nyc, I always have to manually look up their distances to figure out if I can make it to each in time. There's been other times I've been hanging out with friends, and we want to do something in the local area wherever we are but are unaware of what's going on in a specific local hood in any given city.

A personal experience that inspired this is when in Hong Kong I visited the Tian-Tian Buddha and on the way back found out there was a more authentic fishing village nearby- I would have loved to have realized that earlier and gotten to include that experience into my trip. I had a similar experience where I missed a local band in Berlin because I found out about it after I had already left the area I was in. Being able to see what other activities and/or events are nearby can condense transit time both while at home in nyc or abroad to improve time management.

What it does

This app condenses events calendars onto a map to help people better plan their schedules for both at home and travel itineraries. It populates events by categories that a person is interested in based either on their person's geo-coordinates using a location feature to find events near them or let's a person put in locations they will be traveling to in order to simplify the itinerary planning process for trips or nights out. It also regularly updates based on where a person is, their availability and autosuggests events based on their preferences, interests and history.

How we built it

We used Eleven Labs, Lovable, Firecrawl, Supabase, Github, Mapbox, Google Maps to build it.

Challenges we ran into

Loveable wrote the code super-fast and was able to make changes to the back and front ends at lightening speeds. When it would rewrite errors in logic it would wipe out work previously done and it had to be done many times over before we were able to have consistent pages. It also would duplicate buttons and tabs and not erase the first ones which caused a lot of problems in the UI and organizational structure. Lastly, at times it seemed to be hallucinating with output and several times each event calendar had to be refined because it was randomly switching city pages on us between the cities despite predefining the directories and geocoordinates several times, so in some ways it might be easier to hardcode some parts of the database with specific crawling outputs for the events calendars. That said, in other ways, it significantly expedited creation- even though it is not close to done and still has some issues. Also, the Google Maps api kept having to be readded- which is normally not the case- this seems to be a glitch and we'll either use a different mapping api or we'll troubleshoot the glitch-still assessing what would help resolve the problem.

Accomplishments that we're proud of

Setting up the multi-city, multi-category, multi-event crawling feature to update regularly both on static and dynamically based locations. Learning to use two new programs- Lovable and Firecrawl. Even if not complete and still needs a lot of work, having a working demo is still something to smile about.

What we learned

This was the first time working with AI to build an agent and it was a bit of a learning curve. If I did it over, I would be more specific in the hierarchy and pathways from the beginning. It probably would have cut down on the rewrites in the code which caused several challenges in having to repopulate pages. I also might hard code parts of the database from the get-go. The idea of being able to have AI do all that work was so alluring I was eager to see the outcome. All in all, it's pretty good but I should have been more proactive in the backend architecture. I also learned that the AI responds better to screenshots than text sometimes in understanding what the problem is and how to correct it.

What's next for Explorologie

We'll continue to build out the events calendars with more localized content for natives and for people who want to travel and immerse themselves more in the local cultural events of any given city. We'll add on more travel suggestions so people can streamline their daily schedules, event calendars and travel itineraries. We'll improve the mapping stability, localized events options in the calendars and let people buy tickets for ticketed events from us for a fee. We'll sync up with Uber and Lyft so people can order their Ubers in advance based on their schedules and travel itineraries.

We might add a geolocation pinning where people can tag photos and travel tips that can only be accessed by either going to the map or the physical location.

Built With

• amazon-web-services
• elevenlabs
• firecrawl
• google-calendar
• google-maps
• loveable