Building a voice-AI platform for two use cases: proactive workplace safety check-ins that catch hazards before they become incidents, and a voice-native legacy system translator that helps engineers modernize COBOL by reconciling spoken intent against actual code logic. We need a backend/infra dev comfortable with Rust and Go (Go for the SFU/WebRTC layer, Rust for command logic), someone who can work with AssemblyAI’s voice API, and a designer to make the conversational flow feel simple and trustworthy rather than clinical.