I need a Raspberry Pi configured as a personal voice-first assistant. The device must: • Use Vosk for on-device speech recognition • Speak responses through the Piper text-to-speech engine • Store and recall past conversations with reliable long-term memory (a lightweight database or comparable solution) Scope You’ll handle both hardware and software bring-up: flashing the OS, installing dependencies, configuring microphones/speakers, and wiring any required GPIO accessories. Once the core stack is running, create a clean API or service layer so future skills can hook into the speech pipeline. A functional prototype that boots and is ready to converse is the end goal. Deliverables 1. Fully imaged SD card or repeatable install script 2. Source-controlled configuration files and code 3. Brief documentation outlining setup steps, memory schema, and how to extend the assistant I’m comfortable with Python, so keeping the codebase in that ecosystem is preferred, but I’m open to the language that best supports Vosk and Piper. Reliability and smooth, low-latency interaction matter most.