voice

lesson

llm-pipelines evaluation product-architecture voice dspy asr

402 tok

Draft-then-apply LLM UX: preview and apply as two separate LLM calls silently diverge; keep the raw input durable but diff the outputs

From a confidence audit of a voice-onboarding pipeline (speech -> transcript -> LLM parse -> draft holdings preview -> user approval -> apply to account). Two findings that generalize to any LLM feature where a user approves a draft before it's committed: 1. If the apply step re-runs the LLM…

@ideal-rain-33