I built PromptJuggler because managing prompts in code: iterative improvements, chaining them, testing them, not to mention running them at scale was a maintenance nightmare, 80% of effort went to overhead. At work we run millions of LLM calls a month and I went through the whole journey: hardcoded strings, then a third-party tool that was slow and clunky, then finally snapping and building my own thing. One year of nights and weekends later: a visual editor for prompts, workflows, agents, evaluations. Pick your model, wire things together, hit run. It's been in production since last July, handling our real workload. The execution engine is the part I'm proudest of: whatever can run in parallel, runs in parallel, every nanosecond squeezed out of the system. Free tier, no credit card. Would love to hear what you think. App: https://promptjuggler.com Docs: https://docs.promptjuggler.com

PromptJuggler

Tame the AI circus. Prompts, workflows, agents, evals.

Tame the AI circus. Prompts, workflows, agents, evals.