Hily dating app uses AI and identification to improve safety and potential matches
Dating through apps has several ongoing problems. The glaring privacy breaches seem to mount up day by day. For instance, cyber-security researchers realized that hackers could create a creepy Chrome plugin allowing a Tinder user to find the precise location of their Facebook friends who were also on Tinder. A journalist asked Tinder for the data it had on her and got back 800 pages of detailed information which, if hacked, could have exposed her most private life. And last month a young mother killed herself after malicious trolls set up a fake Tinder account in her name.
What if the only way to set up a dating account was to verify an actual identify, preventing fake accounts / catfishing, trolling and improving the overall experience with AI? That’s the aim of a new startup, Hily.
Co-founder Yan Pronin comes from the world of analytics, M&A modeling, dynamic pricing and statistical modeling. He realized he could bring some of that learning to an app which could improve the way people connect and scale it through pre-defined events that are based on mutual interests.
Hily works by employing matchmaking algorithms which are based on machine learning instead of the geographical location of a user or “attractiveness level” (as used by Tinder) which might keep you swiping in the app but actually reduces your chances as around 80% of people are rejected.
Hily ignores attractiveness levels and goes for better matches, identifying users with the same interests and a higher probability of matching, taking data from the depth of dialogue, mutual likes, photos sent etc. The more the more person uses the application, the higher are the quality of his/her matches.
Users don’t get an attractiveness score but a “risk score” depending on the user’s profile (passing verification, complaints, depth of dialogue, activity, etc.). If a person has a high-risk score, the app can block other users from sending their personal information to that particular profile.
More significantly, in today’s strange world, however, is creating the safest environment. So Hily uses ID verification via Facebook, photo capture and (soon) official ID. It also prevents people from uploading pictures which have been tampered with. It also has an “Events” feature, allowing users to meet at a public event, to increase safety.
Everything on the app is free and so far it has 35,000 users in closed beta.
Originally published at techcrunch.com on October 5, 2017.