Dating Apps are Hardcoded to Match Looks. Wavelength's AI Wants to Fix That

The burnout on dating apps is real. Many users, especially GenZs, are fatigued by endless swapping that takes a considerable emotional toll. The fatigue is as high as among 79% of GenZ users, according to a Forbes Health survey.

Many dating apps are now using artificial intelligence to compress swiping. Tinder is using AI to personalise curation on the basis of user interests and behaviour. Hinge offers a conversation assistant, while Bumble’s AI learns users values and relationship goals to make their case to others.

However, Jayanth Emmadi, Co-founder and CTO of dating app Wavelength, believes those modeld reward attractiveness more than compatibility. “Most platforms match people based on how many likes they receive,” he tells AIM.

Wavelength is also trying to rethink online dating and moving away from swipe-based matchmaking, but not with a square focus on looks. Co-founded by CEO Ishaan Gupta, COO Vinit Sarode, and Emmadi, the platform offers an AI companion called Wave that gets to know users through conversations before suggesting potential matches. The idea is to spend time understanding a user's personality, relationship goals, and preferences through natural interactions.

"We wanted people to feel like they were talking to someone who genuinely understands them, not just another chatbot," says Emmadi.

Building Conversations

Unlike traditional dating platforms, where users immediately begin browsing profiles, Wavelength asks users to interact with Wave first.

However, early on, the founders realised that simply creating another chatbot would not solve the problem.

“Building a companion that truly understands you and remembers you was one of our biggest challenges. If a person tells Wave about someone they went on a date with and the next day it completely forgets that conversation, the person will feel irritated because it’s supposed to understand them.” Emmadi notes.

To overcome this, the team designed Wave with long-term conversational memory, enabling it to retain prior interactions and continuously improve its understanding of each user.

But the challenge extended beyond conversations. Most dating platforms rely on variations of the Elo rating system, originally designed to rank chess players. Users who receive more likes are generally shown to similarly popular profiles.

Changing that meant developing an entirely new matchmaking algorithm that combines conversational insights with behavioural patterns rather than popularity scores.

“We understand you first and recommend matches based on that understanding- not just whether people are liking you,” he adds.

Learning From 6,000 Users

Before launching publicly, Bengaluru-based Wavelength, founded in September last year, spent months in beta.

The founders built two WhatsApp communities, one for men and the other for women, with a total of nearly 6,000 members. Users regularly shared feedback, while the founders also conducted one-on-one calls to better understand their experiences.

Since launching publicly in September last year, Wavelength has grown to around 12,000 users in India. Its app has a 4.0 rating on Google Play Store. The founders say that nearly 95–96% of its user acquisition has come organically through social media content and word-of-mouth referrals.

Those conversations revealed something unexpected.

“The first thing we realised was that people start trusting Wave when they feel understood. Sometimes Wave gives users an insight about themselves that they already knew deep down but had never consciously recognised. That’s when trust begins,” says Jayanth.

“Instead of overwhelming users with hundreds of profiles, Wave attempts to simplify decision-making by understanding individuals first,” he explains.

Looking Beyond Looks

One of the team’s biggest surprises came from analysing more than 800 successful matches during beta testing. Contrary to the common perception that dating apps favour conventionally attractive users, many successful matches involved people with compelling personalities or unique life experiences.

“These were regular people who had some crazy spark in their lives. Maybe they had travelled to 26 countries, built something meaningful professionally, or were simply emotionally available,” Jayanth notes.

Rather than highlighting appearance, Wave introduces users through stories.

“Someone who climbed Mount Everest alone may hesitate to write that achievement on a dating profile for fear of appearing boastful. But Wave can naturally explain what that accomplishment says about the person’s resilience and character,” he affirms.

For the founders, helping personalities become visible beyond photographs is one of the platform’s biggest advantages.

Relationship Companion

Interestingly, many users continued interacting with Wave even after finding partners.

One feature, called Debrief, allows users to discuss dates, seek advice after difficult conversations, or simply reflect on their experiences. According to Jayanth, some beta users even replaced general-purpose AI assistants with Wave for relationship-related conversations.

“Some of our friends who found partners through Wavelength still use Wave, not because they want more matches but because it has become like a mutual friend who understands both people,” he explains.

According to findings published by researchers from the Federal University of ABC, Brazil, AI-powered systems have the potential to move dating platforms beyond quantity-driven interactions and towards more meaningful compatibility.

For the founders, that suggests AI could eventually play a larger role than simply introducing people.

Solving Problems

Like any young product, Wavelength has encountered several design challenges.

One early feature automatically expired matches after 24 hours—a decision intended to encourage quicker responses. However, Jayanth acknowledges that many users felt frustrated after missing potential connections simply because they had not opened the app. 

“Following user feedback, the company extended the window to 48 hours and is now exploring more flexible systems where timers begin only after users view recommendations,” he adds.

While the company is currently focused on product development and user growth, its long-term business model will centre on premium subscriptions, with paid features planned as the platform matures.

The startup currently operates with a team of seven people. In December last year, it raised $500,000 in pre-seed funding from Antler after participating in the accelerator’s cohort. It is currently refine the product before it looks for growth capital.

Another challenge lies in understanding rejection.

“People usually say there was ‘no spark’. They don’t want to sound rude, so it’s difficult to collect meaningful feedback. But without that feedback, it’s also difficult to improve recommendations”, he explains.

The team is also working on improving how Wave introduces potential matches. “Instead of simply informing users that someone is ‘99% compatible,’ the AI attempts to explain why that individual might genuinely be worth getting to know,” he notes.

Beyond Dating

While Wavelength began as a matchmaking platform, its founders now envision something much broader. Jayanth believes AI could eventually help people maintain all kinds of relationships, not only romantic ones.

“There will be an AI for your professional life. I think there will also be one for your social life,” he confirms.

He imagines Wave helping users reconnect with old friends, strengthen family relationships, suggest activities for couples, or even mediate conflicts when needed.

As artificial intelligence increasingly moves beyond productivity tools into everyday human experiences, startups like Wavelength are asking a new question, not whether AI can recommend better matches but whether it can help people feel understood before helping them fall in love.