Introducing the new visual workflow builder for AI agents

Building AI agents has traditionally required deep technical expertise. You needed to understand orchestration frameworks, manage API connections, handle error states, and write thousands of lines of glue code. For many teams, this meant months of development time before shipping anything to production.

Today, we're changing that with our new Visual Workflow Builder.

The Visual Workflow Builder lets you design complete agent workflows using a simple drag-and-drop canvas. Connect LLMs like Claude or GPT-4, add tools like web search or database queries, and define logic branches — all without writing infrastructure code. What used to take weeks now takes hours.

We built this because we believe the future of AI development should be accessible to everyone. Product managers should be able to prototype agent ideas. Engineers should spend time on business logic, not boilerplate. And teams should be able to iterate quickly based on real user feedback.

The builder includes 100+ pre-built components covering everything from LLM routing to email sending to document parsing. Each component is production-ready, battle-tested, and designed to work seamlessly with the others. Simply drag them onto the canvas, connect the dots, and deploy.

Every workflow you build comes with built-in observability. You can trace every decision your agent makes, debug issues in real-time, and optimize performance based on actual usage data. No more black boxes.

We're already seeing teams cut their development time by 10x. One customer went from idea to production agent in a single afternoon — something that previously took their team three weeks.

The Visual Workflow Builder is available today for all users. Free tier includes full access to the canvas and 50+ components. Pro users get access to the complete component library and advanced features like custom branching logic and A/B testing.