Abamama _ Azariah team 2020 Dev Hackaton Feb 16 at SF

Abamama offers Customized Recommendation Service for Traveller and Foodie at Hackaton using AI,ML,GPS,Image

Comment 1

Inspiration

We are offering travel lovers and Foodies Customized Recommendation when they arrive at destination or while they're passing through must visit spots using Computer Vision, Machine Learning, Deeply Relevant Recommendations, GPS, Real-Time Location. For Local Business owners and SMEs, we offer a way to boost their business from tourists and connect them to potential customers who are willing to explore personalized and relevant recommendations suggested throught the app.

What it does

The user opens the app, clicks 'where to go?' or just type 'Destination', when they arrive in a new city, Abamama offers customized, real-time, 'only for you' Discounts and Free Vouchers to explore the best restaurants and recommend 'Just for you', 'What to do' or 'Where to Visit'. Recommendations are personalized suggestions based on similar users, so the user won't get just the most popular restaurants, but the ones they are most likely to enjoy. This realtime recommendation paired with discounts creates an optimal experience for the user while more narrowly targeting discounts and incentives from relevant shops and restaurants.

How I built it

1) Recommendation engine: TomTom API, Python, Tensorflow, Scimitar models, Google cloud API, also tried Amaedeus and Canon api.

2) Client app: TomTom API, Reactjs, Canon

3) Backend/Hosting: TomTom API, Google Cloud Functions.

Challenges I ran into

We started Abamama service for people at Hackathon yesterday. Our team came from all over the world, including Korea, Guatemala, and a few locals. Challenges we addressed are from the Main Hackathon, TomTom, Amadeus and Canon.

Accomplishments that I'm proud of

We built demo version of Abamama app and are seriously willing to make it as a real app after the competition. It makes people happy and use time efficiently and to save their money, which boosts local business in visited countries.

What I learned

We met at Dev Week Conference and each of us have been learning Cooperation, Expressing our strong skills and dividing best roles to work effectively in limited time, Communication skills to understand different views and Entrepreneurship to work for one project as if it's our real service in the real world.

What's next for Abamama _ Azariah team 2020 Dev Hackaton Feb 16 at SF

We are expecting to collaborate with 3 Hackathon Sponsors and work with Blockchain Digital ID to offer 'Quick, Convenient, Customized Recommendation Service for Travelers and Foodies including Teenagers'.

Built With

  • also-tried-amaedeus-and-canon-api.-client-app:-tomtom-api
  • amadeus
  • google
  • google-cloud
  • python
  • reactjs-backend/hosting:-tomtom-api
  • recommendation-engine:-tomtom-api
  • scimitar-models
  • tensorflow
  • tomtom
Share this project:

Updates

posted an update —

I worked on the backend machine learning and recommendation engine. I used python based computer vision CNN to identify objects and label them from user submitted photos from mobile and canon api uploads, and location information from the TomTom api. Using those labels I built a recommendation engine using a combination of models including clustering, SVM, and Deep Learning Embedding to apply collaborative filtering from similar users. This allowed us to return personalized and relevant recommendations from similar users that align with the user’s interests and location.

Log in or sign up for Devpost to join the conversation.