E-051-Hajj Bot v1.0

HajjBot is an AI personal assistant & concierge serving needs of the pilgrims, before, while and after the Hajj.

Comment 3

Inspiration

+2 million people, +120 nationalities and more then 200 languages are represented in the Hajj annually, in one location. The big question is how do they communicate efficiently when in need of assistance or in case of an emergency.

The reality is they don't and can not, which has a huge time and opportunity cost on the visiting pilgrims and the concerned authorities and organizers.

Millions of people make the pilgrimage every year and millions more wish of performing the ritual once in their lifetime. For many Hajj is a once in a lifetime opportunity, therefore dedicating time, focus and energy towards the rituals is the main agenda. At the same time the pilgrims require a lot of assistance when in case of emergency or assistance.

While during the hajj an estimated +400,000 emergency and assistance calls are made in the 5 day of hajj. 30% of these calls get directed to the wrong agencies/authorities. The Hajj pilgrimage handles the biggest amount of emergency and assistance call volume in a short period of time.

What if the pilgrims can communicate in their languages and get assistance hassle free and efficiently.

We came up with the idea of building an easy to use multi channel and omni platform deployable solution, which can be utilized through many emergency and critical use cases.

What it does

-helps Hajj pilgrims, with assistance on their pilgrimage journeys -helps pilgrims in emergency situations -helps save lives in emergencies. -helps to get in touch with the right emergency department when in an emergency. -provides valuable information related to the query in multiple languages.

How we built it

  • Azure Bot framework
  • Social media platform APIs
  • multiple open source APIs
  • Google Maps API
  • Bing Search API
  • Bing Speech API
  • Bing Cognitive API Services
  • C#.Net/ASP
  • LUIS- Language Understanding API (A machine learning-based service to build natural language into apps, bots, and IoT devices. Quickly create enterprise-ready, custom models that continuously improve.)

Challenges we ran into

-Building solid NLP multilingual intends with LUIS. -Using spell correction in the backend engine. -Using third party AI services (Speech to text, speech recognition, image to text, image search, text to speech, GIS services, image processing) -Deciding on backend platform options. -Deciding on what initial languages to build MVP on -trying to stay calm when a feature got stuck(unable to make it work) for more then 4 hours.

Accomplishments that we are proud of

-Creating a viable business and monetization around the conceptual idea -Getting through all the issues we faced through the development -Deciding on best use cases. -Deciding on best channel delivery options

What we learned

  • Problem solving in tight time-frames
  • Working under pressure
  • Team (management, collaboration, leadership, ...etc)
  • Time Management
  • Exchange of experiences
  • Communicate with people from different countries and dialects.
  • Work on a dream in a planned way (Good Business Plane/Development)

What's next for Hajj botv1.0

  • Implementing NLP Functions into the architecture.
  • Implementing Artificial and Machine learning algorithmic learning to the data sets generated down the road-map.
  • Using more powerful services like google maps API, google search API, google vision API .
  • Use GIS technologies and services from multiple service provider such : ESRI and HERE maps.
  • Use integrations with the government services like Police, Ambulance, Firefighters, Transport, Tourism and Finance ....etc
  • Adding languages support for other speakers by the percentage of attendance.
  • Adding multi lingual speech control and speech recognition capabilities.
  • Integrating third party integrations such as hotel booking platforms, airline booking, hajj data.
  • Implementing Machine learning to build out models defining user and pilgrim preferences.
  • Integrating Hajj IOT infrastructure database services to recommend on traffic, crowding and timing notifications

Built With

Share this project:

Updates