For the past 18 years, we've been deeply involved in developing multimedia projects and crafting products for a diverse range of clients and industries. With our extensive experience in this field, we felt it was time to go beyond client projects and develop our app. We wanted to showcase our expertise by creating something truly exceptional. And thus, BlaBlaPlay was born.

What is BlaBlaPlay?

When we started to think about our the future app, we had the idea of creating a social network like Tinder or ChatRoulette. Our main focus was on chatting, where users could share their ideas, opinions, and connect with like-minded people – so we developed apps both for iOS and Android.

There are 3 types of users we have in BlaBlaPlay: active chatters, observers, and modest ones
Types of BlaBlaPlay users

To encourage active participation, we made the platform anonymous and added a Tinder-like "swipe" feature. In the feed, users can view cards with voice messages. Unlike ChatRoulette, where communication is done via video, in BlaBlaPlay we proposed to use voice only, which makes it easier to communicate on any topic.

When users come across a voice card they like, they can record a response instantly, initiating a conversation with the card's author via voice or text. If they're not interested, they can swipe left to see the next card.

We added a Tinder-like "swipe" feature: swipe right or left to like or dislike the voice card
The Tinder-like "swipe" feature

Initially, we introduced an "energy" scale to encourage users to engage by swiping cards. Each swipe costs one unit out of ten, which is replenished hourly. Users could also contribute to the feed by posting their cards, restoring one unit of energy at a time. We later increased the energy limit to 100 for more communication options. However, we eventually phased out the energy system and introduced a like function for cards instead.

We decided to forego the skip function and introduce the like feature instead
"Energy" scale was designed to encourage user engagement by swiping cards

Complete anonymity is the key feature of the project

Unlike other platforms, BlaBlaPlay prioritizes user privacy by not requiring any personal data. None whatsoever.

When signing up, users only need their Apple ID or Google account. The app then generates various nickname options, ranging from "Batman" to "Musician" or "Slim Shady." With over 7,000 unique avatar combinations, users can select a "chattering" hand with their preferred accessory or background, ensuring there's something for everyone.

Users can generate their own custom avatars and nicknames
Nickname and avatar generation

Content moderation

To ensure a safe environment, we've integrated a content report system into BlaBlaPlay. Any cards containing offensive or inappropriate content are promptly removed by our administrators. Additionally, if a user accumulates more than three complaints, their account will be permanently deleted to uphold community standards.

We introduced a content report system to uphold community standards
The BlaBlaPlay content report system

AI implementation

As BlaBlaPlay progressed, we aimed to enhance user engagement by integrating artificial intelligence (AI) into our app. Our initial challenge was users responding to voice cards with silence, prompting us to seek ways to foster interaction.

Initially, we considered implementing a chatbot to provide predefined responses. However, we opted for a more seamless approach using AI to suggest conversation topics based on voice card content. This led to the creation of AI prompts, where the app translates voice messages into text, processes them through ChatGPT, and displays suggested topics for users to discuss.

We used ChatGPT to suggest conversation topics based on voice card content
AI-generated discussion topics

Another aspect of AI implementation involved moderating AI-generated content. We trained a neural network to identify offensive language in voice cards by translating them into text using CoreML technology. The neural network then flags any prohibited content for administrator review, allowing them to take appropriate action: delete the card or block the user.

We trained a neural network to identify offensive language in voice cards using CoreML
AI-powered content moderation

We've also added smart recommendations. The most popular and fresh content is displayed in users' feeds more often than voice cards that have been reported frequently. This motivates users not only to record voice cards more actively, but also to follow the app's rules. 

To sum up

In just a few months, BlaBlaPlay has transformed from a cool concept into a fully-fledged app complete with unique features, effective content management tools, and engaging user interaction mechanisms. Now we continue to work on further enhancing BlaBlaPlay and welcome new users and feedback.  

Looking to develop your own mobile app? Contact us or book a quick call

We'll discuss your project, brainstorm ideas, and offer you an initial estimate. It’s free.

  • Cases