Hello ProductHunt 👋 - really excited to be launching today! 🚀 😁
First of all, I know… Another (AI) observability tool... 🫠 To be honest, we wish we didn’t have to add to the pile, but in our view AI observability is broken, and it needs fixing.
When building AI apps ourselves, we were often left fighting against the observability tooling, trying to strip things back to the bare minimum.
Unify interfaces addresses this, by being very: 🪶 lightweight 🔧 hackable ⚡fast 🧱composable
We hope that it makes it easier to go from zero to one on your products (with or without LLMs), letting you focus on the metrics that matter to you, and nothing else 🎯
The core building block is simple, just `unify.log`. This lets you store any kind of data to your console for easy visualization, grouping, sorting, and plotting etc.
You can then hack together your own custom interface, for whatever you want using three basic tile types: Tables 🔢, Views 🔍 and Plots 📊
You can use these three primitives to:
➕create + visualize your datasets in a new tab (with orwithout LLMs)
➕monitor and probe production traffic in a new tab (with orwithout LLMs)
➕start an evaluation flywheel in a new tab (with orwithout LLMs)
📉optimize your product for your users (with orwithout LLMs)
🧠whatever else you can think of (with orwithout LLMs!)
We’ve created a case study (with LLMs 😁) showing how you can compose these basic concepts to get clear view of the data + metrics that matters to you (and nothing else 🧘), and ship at lightning speed ⚡
Feel free to check it out and let us know what you think. We’d love all kinds of feedback (especially criticism and suggestions)!
Congrats on shipping Unify! Love the "Notion for AI observability" concept simple, focused, and exactly what teams need to track meaningful metrics without the clutter.
I love how Unify interfaces is stripping down AI observability to just what’s essential—lightweight, customizable, and fast! How do you see teams using it to streamline the data visualization and monitoring process in real-time?
Just tried Unify and it's exactly what i needed for my AI projects! been struggling with other observability tools that feel so bloated and slow. this is refreshingly lightweight and actually lets me focus on the metrics that matter instead of drowning in unnecessary data.
love how hackable the interface is - took me like 10 minutes to set up custom visualizations for my specific use case (which would've been impossible with langsmith or weights & biases without major workarounds).
the ability to quickly compose tables, views and plots is super practical. no more fighting with the tools just to see what's happening in my models!
congrats to the team for solving a real pain point. curious about your roadmap for supporting more complex evaluation workflows? (. ❛ ᴗ ❛.)