Banyan

Hey Product Hunt! 👋 Over the past few months, we’ve been building with LLMs and noticed something… Prompt engineering gets messy fast. We were hardcoding prompts in YAML files. Changes were invisible. Testing was manual. Output quality was hard to measure. And version control? Nonexistent. So we built Banyan — a complete prompt infrastructure platform for teams shipping AI features. With Banyan, you can: ‣ Design multi-step prompt workflows visually, without hardcoding logic ‣ Version your prompts Git-style ‣ Collaborate with your team on a single source of truth for all prompts ‣ Run A/B tests across prompt versions and measure what works ‣ Evaluate prompts automatically using LLM-powered scoring functions ‣ Integrate using our SDKs and CLI for OpenAI, Claude, and more We built Banyan for devs and teams who want to treat prompts like real software — testable, trackable, and improvable. We're in open beta (and free!). You can get started in 5 mins: https://usebanyan.com/ We’d love your feedback, ideas, or questions — and thanks for checking us out! –The Banyan team 🌳