today, we're announcing our fundrai--
jk. we're shipping.
meet raindrop experiments 🧪
00:00
- today, we're announcing our fundrai-- jk. we're shipping. meet raindrop experiments 🧪00:00
raindrop experiments is the first A/B testing suite built for AI agents. it uses your existing feature flags (e.g. PostHog or Statsig) and gives you comparisons built specifically for AI agents. no feature flags? no problem. just compare yesterday to today.imagine you add a new tool (or fix an existing one). use experiments to see how your change affected your agent's behavior there's a breakdown of each tool: % of events it is used for, how many times it gets called per event, what percentage of invocations have errors.click any comparison and dive into the actual user events behind the aggregated data. and send to your team with easily-shareable links. that's experiments on raindrop. it's available on pro plan at
very nice Ben, it would be good to see an example end to end to understand the whole process
