Launching today
Versia.dev
Experimentation Platform for Agents and Teams
9 followers
Experimentation Platform for Agents and Teams
9 followers
AI can produce endless variations, but which ones actually work? Versia is reward-driven feature flags. Define your variants, send a reward when users convert (a click, a purchase, a signup), and Versia automatically shifts traffic to the best-performing option. OpenFeature-native: works with any OpenFeature SDK (Go, Node, Python, Java,NET, and more). No proprietary SDK to install, no vendor lock-in. Get started in minutes!
Hi ProductHunt,
We wanted something that can optimize live traffic continuously based on open-ended rewards (clicks, signups, etc) and context (location, user tier, etc). We also made sure to build it on the CNCF OpenFeature standard.
This enables interesting usecases because agents can close the loop:
generate new variants, autonomously register them as experiments, receive rewards, and promote the winning variants.
A couple of things I'd love your take on:
What's the last thing you shipped where you never found out if it was the right call?
Did the LLM instructions on our site one-shot your first experiment in your codebase?
Let me know!
Interesting take, and easy to set up. Let’s see how it handles some traffic now! Best of luck