Tech Lead8club
A problem you can't solve anywhere else, because nobody else has the data.
At 8club, every accepted invite writes a real edge into a social graph, two people, a night out, a connection that actually happened offline. Not a follow. Not a like. A verified fact about human trust.
So here's the problem we live inside:
An invite from someone you know converts at 44%. From a stranger, 5.6%. That 8x gap isn't a growth stat, it's a property of the graph. Which means the interesting question isn't "who should we notify," it's: given a person, a night, and a city, can you find the single edge in the network most likely to result in two people actually being in the same room?
That edge decays, trust weakens with distance, saturates with overuse, and shifts by context. The same invite that converts on Friday dies on Tuesday. And it has to resolve in under 100ms, at city scale, without a black-box model you can't reason about.
Facebook solved this for a feed. Uber solved it for supply and demand. Nobody has solved it for real human trust, in the physical world.
We're hiring a Tech Lead to own that architecture.
Bangalore, in-person.
And yes, we pay above market. We'd rather spend the money on people than on ads.
Email me at kunal@8club.co or kaushal@8club.co, or tag someone who'd actually enjoy this problem.
tal saysown the architecture for a real-world social graph and trust network at city scale
engineering
lead
in-office
Bangalore
comments 0 anon