Tech, ExplainedGaming, streaming and media
Product

Dating app

Swipe, call or play to meet someone.

BEFORE ITAFTER ITDating appFRNDHunchCOVERSVoice roomsMatchingPaid extrasSelfie checksIdentity verification (IDV)↗ KNOW YOUR CUSTOMERverifiesthe selfieAge estimation↗ KNOW YOUR CUSTOMERestimates theuser's agechecksprofilephotosDeepfake detectionsendsreportedaccountsTrust andsafety investigationCONCEPTS IT USESRecommendation systemsAI-generated contentBEFORE ITAFTER ITDating appFRNDHunchCOVERSVoice roomsMatchingPaid extrasSelfie checksIdentityverification(IDV)↗ KNOWYOUR CUSTOMERverifiesthe selfieAge estimation↗ KNOWYOUR CUSTOMERestimates theuser's agechecksprofilephotosDeepfakedetectionsendsreportedaccountsTrust and safetyinvestigationCONCEPTS IT USESRecommendation systemsAI-generated contentBEFORE ITIdentity verification (IDV)↗ KNOW YOUR CUSTOMER↓ verifies the selfieAge estimation↗ KNOW YOUR CUSTOMER↓ estimates the user's ageDating appFRNDHunchVoice roomsMatchingPaid extrasSelfie checksAFTER IT↓ checks profile photosDeepfake detection↓ sends reported accountsTrust and safety investigationCONCEPTS IT USESRecommendation systemsAI-generated content
The dating app receives verified selfies from identity verification and age estimates from age estimation, and sends profile photos to deepfake detection and users' reports to trust and safety investigation.
opens a page inside this one a page in the same scene a page in another scene a concept page a label only, with no page

A dating app is an app that helps people find a friend or a date, by swiping through profiles, by voice calls, by social games or by answering polls. Most dating apps are free to join, and earn from paid extras. Some are built for friendship as much as for dating.

The best-known way is swiping. The app shows one profile at a time, with photos. The user swipes one way to show interest and the other way to pass. When two people both show interest, it is a match, and they can chat. Swiping means judging someone mainly from a photo. Several Indian apps work differently. On some, people meet through voice calls and games first, and see photos later. On others, people answer polls, and the app matches people whose answers are alike.

Consider a software engineer who has just moved to Bhubaneswar for a new job. He knows nobody in the city. He joins a voice-first app and lists his interests: football, Odia films and cooking. The app puts him in a group voice room with six other people who like football. They play a quiz game together. Afterwards, he and one person from the room start a one-to-one voice call. Only after a few calls do they choose to share photos.

Who is shown to whom? A recommendation system decides. It learns from each user's choices: whom they call back, whom they skip and how long their calls last. Then it ranks the people it shows next.

The app earns from paid extras. A user may pay to see who liked them, to appear higher in other people's lists, or to make more calls in a day. Many apps also sell coins, which users spend on gifts during calls. This is explained in virtual currency.

Safety is a large part of the work. When someone joins, the app asks for a selfie and checks that a real person took it. Identity verification (IDV) matches the selfie with an ID document. Age estimation estimates the user's age from the face, to keep out anyone under 18. Profile photos can be made with AI, so the app may also use deepfake detection. And when a user reports an account, for example for asking for money, the report goes to trust and safety investigation.

AI is now used on both sides of the app. It helps users write their profiles and suggests a first message. It is also used by scammers, to create profiles that look real. This is covered in AI-generated content.

Tech, Explained is a part of ingrid.fyi. Every page explains one piece of technology in plain words.