I want to turn my laptop into a hands-free, voice-activated companion that feels truly alive. The assistant should wake on a custom hot-word, listen continuously, and respond in natural speech with near-zero latency. Voice commands are the top priority; text is secondary and can be added later. Core functionality I need right now: • Wake-word detection and always-on listening that consumes minimal resources • Instant, conversational replies sourced from an ever-expanding knowledge base (GPT-4, vector search, or similar) • Mood awareness: the system must sense when I sound happy or sad and adjust tone, pacing, and word choice on the fly. If you can future-proof it for other emotions like anger, even better. • Full PC control: browsing and managing files, launching or killing apps, plus system diagnostics and updates—essentially anything I could do with mouse and keyboard. • Continuous learning so it refines responses, remembers preferences, and improves predictions day by day without manual retraining. Acceptance criteria 1. Saying the wake word triggers an audible acknowledgment within 300 ms. 2. A verbal prompt such as “Open today’s report and email it” executes flawlessly without touch input. 3. If I greet it sounding upbeat versus downcast, its reply clearly changes tone while the transcript logs the detected state. 4. Knowledge retrieval must cite or link sources when asked “How do you know that?” 5. All data stays local unless I explicitly grant cloud sync. You may choose the underlying stack—Python, Node.js, Rust, Whisper, Rasa, LangChain, or any blend—so long as the deliverables run on Windows 11 and can be re-compiled or re-trained by me later. I’ll supply API keys and test scripts; you package the code, installer, and a short setup guide.