The biggest technical challenge we faced with Caret was determining whether LLM-generated workflows actually worked. Here's what we learned and how we're approaching it differently next time.
From pair-programming in IDEs to coordinating teams of autonomous agents. Here's how coding agents are evolving and what comes next.
We run up to five coding agents simultaneously, and they often one-shot solutions. Here are our secrets to maximizing their value.
How recording a Loom video and converting it to a GitHub issue with AI creates better context for coding agents.
We use AI to automatically map arbitrary webhook payloads or emails to workflow inputs in Caret.
Discover how we use AI to bridge the gap between natural user input and complex API requirements, making integrations intuitive and error-free.