Global Vote


Project Background

Global Vote is a search platform, where people can search for businesses around them and across the globe based on humanistic opinion and experience. The viewpoint of the system is based on the simple concept of business ratings and reviews. Here, Every vote to a business is made by an actual legit independent human, a customer/user/expert/professional, and not by a robot or a computer program, so every vote can be trustable and as votes and reviews are trustworthy and valid, the search function of the system utilizes these as a parameter to rank businesses. Better the rank, higher the business on the search list.


We were required to develop a search engine using Hyperlocal list and map view, which examines businesses from a registered global vote system on the basis of proprietary machine learning bucket algorithm that considers multiple factors while ranking businesses.

  • 1. Dashboard with Business/agent/Accountant/Admin login to access diverse data which is required by respective users.

  • 2. Agent, being the main component of the system is responsible for adding and validating the businesses to the system. Agent hierarchy is defined which comprises of a tree-like structure for MLM marketing which works as a catalyst for Agent’s growth.

  • 3. Payment gateway through which Accountants can make payments to agents after checking the amount of work that respective agents put in.

  • 4. Users can present rating and reviews to the respective business with a unique phone number related to their account.

Technical Stack

  • Python
  • Javascript
  • Angular-5
  • MySQL
  • Elastic Search
  • AWS
  • Android
  • IOS
  • Google maps
  • Celery

Project Screens


  • 1. We used elasticsearch to generate results but that wasn't enough. We had to write a layer on the top of the elasticsearch which intends to give the list of well-sorted businesses according to business logic and that becomes a proprietary algorithm of Global Vote.

  • 2. With the help of google maps, we pinpoint all the businesses on the map with a critical radius around the user. To make this system scalable, we designed an algorithm which results in providing us top businesses along with good user experience within a particular range, considering the expanded level too.

  • 3. By using a layered architecture and cluster-based design, we easily scaled up the system horizontally. Currently, the developed system is capable of handling up to 1 million users and 10 million businesses without any performance issues.


  • 1. The main challenge we faced after starting the project was to implement a custom algorithm for the business logic and to present businesses on the map which would help a user see the physical location of business easily. the main idea behind Global Vote. By solely using elasticsearch, it would not take us anywhere, as it only gives us a score when we tend to search for a keyword in the system.

  • 2. To have a load on which 100 thousand data points can run smoothly. We had to target at least 10 million businesses and 1 million users in the first year itself. Result speed with satisfying search result was required so it won’t impact the brand value of the business.


We have successfully launched the first version of Global vote with 1 million users and 10 million businesses with quick response time. The search engine works smoothly with different filters and map search is uniform.

In the second period, we have launched a responsive web application along with android and ios apps.
Currently we are working on advertisements, real-time chat, email, and multi-level agent marketing model.

Get In Touch

Let’s meet and discuss your idea over a cup of filter coffee and we’ll help you make it the next big thing of the decade