Friendspire – Case Study

About the client: An Overview

Friendspire is your best friend forever when it comes to finding what to watch, read or eat. This robust app allows you to find great movies, series, books, podcasts, restaurants, and bars based on suggestions given by your friends, family, and loved ones. We do know that there are thousands of other apps that serve the same purpose, then why Friendspire?

If you are looking for a solution that could provide you a complete personalized experience with lists that are curated by your interests, liking, and rating, then Friendspire is your answer. Rather than providing you recommendations based on user reviews, this trailblazing app filters all the reviews based on similar user preferences and behavior to deliver you the most tailored experience. On top of all, Friendspire has curated lists you can look at made by anyone, even celebrities.

From suggesting you the books that you can't keep down to restaurants that would make your night, Friendspire has all sorted out for you.

Achievements !

1M Reviews

Friendspire recently crossed one million reviews, thanks to their awesome community for sharing the best recommendations and spreading good vibes.

1500Users

The First 1,500 users were acquired in 6 months; the second 1,500 new users were acquired in one day.

100K Users

Recently crossed 100k users – a scintillating social recommendation application developed by Debut Infotech has reached a milestone of 100,000 app users.

A look at the visuals of the solution- Friendspire app

Connect with this amazing social recommendation app

Building a social recommendation app to help users find and save their favorites

Two years back, Friendspire approached Debut Infotech with a request to develop a recommendation application that could help users to cut through the clutter and easily find their favorite movies, shows, restaurants, bars, etc. with the help of their friends' suggestions.

Although the app received a lot of traction with the number of downloads crossing the 100k mark in no time, the client was well aware of the limitations of the application that restricted them to add certain features. Based upon the reviews and feedbacks provided by users, investors, and stakeholders, one of the biggest challenges the client was facing revolved around the scenario, “What if your friend is not registered with Friendspire?” Because the social feed was a mirror of likes, dislikes, and browsing patterns of your friends, the absence of such users was the biggest dilemma the client was facing.

Placing their trust in our one of its kind app development solutions provided by Debut, the client wanted to add additional functionality to overcome this issue.

Debut’s AI Solution – The journey from an idea to a multi-functionality recommendation app

A successful solution comes from hours of planning and implementation. Debut Infotech selected the best set of app developers to help the client build an app that would stand out in the crowd.

Integration of AI/ML came up as the biggest solution to track the browsing pattern of like-minded to provide a better recommendation to people who don’t have any registered friends on the app. Having said that, bringing up new functionalities involves some risks. A simple error in the system can lead to a huge loss for the business.

Debut Infotech’s talented pool of developers came up with a solution which when went live on android and iOS platforms, received a great response from the users. In no time, the Friendspire app crossed hundreds and thousands of downloads and brought laurels to our client.

Debut Infotech helped fulfill the client demand by integrating the proposed system in the following conducts:

1
Integration of AI/ML based systems that would help get recommendations from users who were not previously registered with the app, to track their browsing patterns on Friendspire, and provide recommendations based on this knowledge.
2
Provide personalized recommendations regarding TV shows, movies, books, podcasts, and restaurants.
3
A custom-made model was deployed to rank items based on hotness rating. Log scale vote accumulation is done in order to determine the preferences of users who are not friends but possess similar browsing patterns.
4
For Prototype API development, flask-based code was deployed to obtain inference from models. Postman exports were cast-off to validate all APIs.
5
Mobile apps with real-time recommendations were delivered through API integration for getting recommendations and submission of interaction data.

A logistics regression model was used to build a combination of latent factor and neighborhood-based models to rank items in the apps.

Technologies we use to Develop AI Solutions

If you also want to move out of the box and achieve similar milestones in your app
development journey, visit us now!