M-074-Subul سُبُل

• 

  Ability to embed quick surveys in the App

• 

  Finding "trusted" volunteer service providers around you.

• 

  Pilgrims can report incidents or complaints, enriched with multimedia or GPS information

• 

  Pilgrims can report incidents or complaints, enriched with multimedia or GPS information

• 

  Official "targeted" and "multilingual" notifications

• 

  Pilgrims posting their experience in Hajj and giving tips to other pilgrims

• 

  Logo

Inspiration

• 40% of pilgrims are NOT tech savvy. We want a solution that provide benefits not just to the hajj that has it, but also if a nearby person has it (or if someone could originally have installed it for the Hajj!).

• Hajj data is lost every season, and whatever left is scattered all around on various social media platforms. We want a solution that allow for mass-collection of data for post-facto studies of the Hajj event.

• The project has to be commercially viable; that is it serves Hajj and Umrah as well as being applicable/scalable to other industries too.

What it does

The front-end of the solution is a social media application that facilitates communication and mutual support between pilgrims. It also facilitate targeted announcements from Hajj officials to pilgrims.

The back-end uses data (collected from the front-end application or elsewhere) to provide real-time, batch processing, and post-facto insights on Hajj.

How I built it

The data collection part is built using MemSQL database and SPARK technologies.

The analytics models (e.g. sentiment analysis and toxicity prediction) are built using Machine/Deep Learning technology.

The dashboards are built using PHP, HTML, and CSS.

The front-end mobile application and dashboards are built using ionic 3, angular js, type script, and node js.

Challenges I ran into

Data .. Data .. Data ! The unavailability and unreliability of data was a major obstacle in building the analytical ML models.

Accomplishments that I'm proud of

The big data environment works in near real-time; maximum 2 seconds to process and label any incoming post (posts arrive at a rate of 10,000,000 per day) and maximum 2 seconds to answer a query that requires a scanning of 1 Billion posts.

What I learned

Passion and Consistency beats Talent every single day, and twice on the weekend!

What's next for Subul "سُبُل"

We are now working on the front-end mobile application, and intend to provide a working version during this Hackathon.