When you plan a trip, no matter if it's a long or a short one, we sometimes tend to forget to bring some items with us. We've talked to other TechCrunch participants and sponsors, and we've found that most of us (7 out of 10) have forgotten at least 1 item, weather it was a towel, a tooth brush or an extra pair of socks. There are no Alexa skill yet, to help humans doing their baggage.
Don't forget any more necessary stuff when you travel no matter your destination, time on the year or accommodation preference.
What it does
Helps you pack for your next trip, based on location, weather and accommodation preference. (hotel, hostal, campsite)
How we built it
We first identified how Alexa could give us the most accurate suggestions to help us do our baggage by interviewing people and our selves. So we have created a skill for Alexa, called "TripComing" to be able to help you pack your baggage. Using aws lamda functions cloud service, nodejs, mongodb and external weather channel api from IBM Watson, to combine them along with the most common and important items you could need for your trip.
Challenges we ran into
- Allowing Alexa use our own server, it requires https and was not able to get the setup done in less than 24hs. (It would have been way better for developer experience to be able to use an http server in dev mode).
- Alexa is like a little kid, in the sense we had to explicitly tell her variations of the same question which had the same meaning to help her give the right answer.
- Change our mindset. In most of our tech products, we are used to UI interfaces. Thinking about a different interface (and such being not human of course), voice interaction is very challenging. For users and developers.
Accomplishments that we're proud of
Alexa is now able for the first time in history, to help anyone pack their baggage. And this first prototype, opened us a huge set of ideas to create a remarkable product that could go beyond this hackathon.
What we learned
We learned that human-like interaction is humongously complex. But we have had similar feelings in the past (great scientists did) with areas that we now take for granted. Even in 24hs we can make a difference and push the unknown, a little bit into the known. For a good cause.
What's next for TripComing
Improve suggestions engine using previous user data, creating better data models and with more integrations (specially with businesses). Be up to date for whenever payments are integrated into Alexa, as a huge enabler for making a better service.