Finding real discounts online is surprisingly difficult — too many sites recycle expired codes or fake deals.
With Discountime, we’re experimenting with a hybrid approach:
– AI aggregation: crawling and classifying discount codes across multiple sources.
– Human verification: community + staff double-checking codes to ensure they actually work.
The goal is to solve the “trust problem” in coupon/deals platforms.
I’m curious:
– Has anyone here tried building something similar (AI + human curation)?
– How do you balance automation with trust in crowdsourced data?
– Any ideas on keeping the model sustainable while avoiding spammy UX?
Would love feedback from folks who have worked on recommendation, aggregation, or “human-in-the-loop” systems.
RobertChin•2h ago
With Discountime, we’re experimenting with a hybrid approach: – AI aggregation: crawling and classifying discount codes across multiple sources. – Human verification: community + staff double-checking codes to ensure they actually work.
The goal is to solve the “trust problem” in coupon/deals platforms.
I’m curious: – Has anyone here tried building something similar (AI + human curation)? – How do you balance automation with trust in crowdsourced data? – Any ideas on keeping the model sustainable while avoiding spammy UX?
Would love feedback from folks who have worked on recommendation, aggregation, or “human-in-the-loop” systems.