Hi Product Hunt!👋 I’m Furkat Kasimov, the maker behind LAMF (Liquid Audience Measurement Framework). We have all been on the receiving end of dumb marketing automation. You view a product once, and suddenly you are bombarded with emails and push notifications. The problem is that traditional segmentation groups people into slow-moving buckets, and standard event triggers fire blindly the second you take an action. Neither respects the fact that user intent fades quickly. I built LAMF to fix this. It is an open-source framework that treats user intent as a "segment-of-one" state that naturally decays over time. Instead of just blasting messages, it uses a governed policy (like a capped contextual bandit) to decide the next best action while respecting frequency caps, latency, and trust costs. Most importantly, it knows when the best action is simply to **do nothing**. I recently wrote an academic paper on this system. In offline replays using real e-commerce datasets (RetailRocket and REES46), governed liquid policies matched the conversion proxies of always-on triggers while dropping off-category, annoying messages to zero. The Ask: The code and offline benchmarks are live on GitHub, but offline replays are only the first step. I am looking for 2-3 mid-sized e-commerce, SaaS, or product teams to pilot LAMF in a live production environment. If you have an active subscriber base and want to improve your conversion-to-trust tradeoff, I would love to partner with you to test this in the wild. I will be hanging out in the comments all day! What do you think about how we handle user intent today? Let me know if you have any questions about the framework, the math, or the implementation.

Open-Source Framework to Stop Spamming

Increase Conversions and Reduce Unsubscribes

Increase Conversions and Reduce Unsubscribes