Lead AI Architect – Voice-First Hybrid AI Guide for Wearable Wellness Platform

Lead AI Architect – Voice-First Hybrid AI Guide for Wearable Wellness Platform About Us VitalSines International is developing the Inner Intelligence Armband, an advanced wearable platform designed to help people improve wellness, fitness and longevity. Unlike most wearables, our vision is not simply to collect physiological data. We’re creating a voice-first AI Guide that develops a long-term relationship with each user while helping them understand their body through biosensing. Initially the armband will measure: * Motion-tolerant SpO₂ * PPG * Heart Rate * Heart Rate Variability * IMU motion * Contact pressure between the optical module and the skin Future versions will add ECG, blood pressure estimation, respiration, camera-based motion capture and additional physiological sensing. Our goal is to create an AI Guide that users genuinely enjoy spending time with over many years. ⸻ The Project We’re looking for a senior AI engineer/architect to design and prototype a hybrid AI Guide. The Guide should operate using both: * Local AI on a mobile device * Cloud-based AI The user should experience one seamless Guide regardless of where inference occurs. Immediate interactions should occur locally whenever practical. The cloud should provide deeper reasoning, long-term memory, personalization and future multimodal capabilities. ⸻ Example User Experience A new user opens the app. Instead of reading setup instructions… a friendly Guide begins a natural conversation. The Guide helps the user position the armband. As the user tightens the elastic band, the Guide receives: * pressure sensor data * PPG quality * IMU motion * signal quality The Guide says things like: “A little tighter…” “That’s much better.” “One more small adjustment…” “Perfect.” The Guide then remembers that person’s optimal pressure. In future sessions, only pressure feedback is needed to reproduce the same fit. This calibration becomes the first conversation in a relationship that may continue for years. ⸻ Phase 1 Objectives Your first responsibility is not building the complete product. Instead, we want your help designing the overall architecture and creating a working proof of concept. Deliverables should include: System Architecture Design a complete architecture covering: * wearable firmware interface * BLE communications * local mobile AI * cloud AI * long-term memory * synchronization * offline operation * security * privacy ⸻ Mobile AI Recommend and prototype an on-device AI capable of: * voice conversation * pressure calibration * BLE communication * local reasoning * maintaining conversational continuity * operating without internet when practical ⸻ Cloud AI Design cloud components for: * richer conversations * longitudinal understanding * personalization * memory * multimodal reasoning * future camera-based movement analysis ⸻ Prototype Build a working demonstration that includes: * BLE communication (real or simulated) * voice interaction * local AI * cloud AI * pressure calibration workflow * memory * configuration management ⸻ Skills Required We are looking for demonstrated experience in: * LLM engineering * AI agents * Hybrid local/cloud AI * Voice AI * Mobile AI * Swift and/or Kotlin * Python * BLE * Wearable devices * Real-time sensor processing * LangGraph or equivalent orchestration * Retrieval-Augmented Generation (RAG) * Memory architectures * FastAPI or similar backend technologies Experience deploying AI directly on mobile devices is especially important. ⸻ Nice to Have Experience with: * Health technology * Fitness technology * Biosensors * Apple Health * Google Health Connect * Camera-based pose estimation * MediaPipe * OpenAI APIs * Local language models * TensorFlow Lite * ONNX Runtime * llama.cpp * ExecuTorch * Core ML ⸻ What Makes This Project Different We are not trying to build another chatbot. We’re creating an AI Guide that users should genuinely enjoy spending time with. The Guide should become: * coach * teacher * encourager * observer * companion in healthy living The relationship should mature naturally over months and years. The technology exists to build this. The challenge is combining: * physiological sensing * local intelligence * cloud intelligence * memory * conversation * personality into one seamless experience. ⸻ We’d Love to See Please include examples of projects involving: * AI agents * Mobile AI * Voice AI * Wearable devices * BLE * Local AI inference * Hybrid edge/cloud AI * Long-term conversational memory Links to GitHub, publications or demonstrations are appreciated. ⸻ Please Answer These Questions 1. Have you deployed AI directly on iOS or Android? 2. Which mobile inference frameworks have you used? 3. How would you divide this project between local AI and cloud AI? 4. How would you maintain one continuous personality while switching between local and cloud inference? 5. How would you architect memory so the Guide gradually gets to know the user while respecting privacy? 6. Have you worked with BLE wearable devices? 7. Have you built streaming voice AI? 8. If selected, what would your first 30-day development plan look like? 9. How would cybersecurity be secured ⸻ Long-Term Opportunity This is expected to become a long-term engagement. Successful completion of Phase 1 may lead to responsibility for development of the complete AI ecosystem supporting the Inner Intelligence platform.