/ Apr 21, 2026
Open Source AI · Developed 2025

Senior Health Assistant

An AI-powered web application built to help elderly individuals log fitness activities, monitor health metrics, and receive personalized wellness guidance — running privately on your own infrastructure.

100%Open Source
LocalAI · No Cloud
VoiceEnabled
MultiUser

LifeTech’s mission is to enhance senior adult health, wellness, and independence. The Senior Health Assistant is a direct embodiment of that mission — a fully deployed AI platform built in 2025 that puts a personal fitness coach in the hands of every elder adult.

— John Huang, Founder, LifeTech Inc.

Senior Health Assistant Capabilities

🏃
Fitness Activity Logging

Users log walks, swimming, jogging, and more — with duration, distance, and date. The AI tracks trends over time and surfaces personalized insights at every session.

AI Health Analysis & Coaching

A custom-trained Senior Fitness Coach model analyzes health metrics and Apple Health data exports, generating personalized recommendations, weekly summaries, and wellness reports.

🎙️
Voice-Enabled Interaction

Full speech-to-text and text-to-speech support with configurable voice profiles. Designed to minimize barriers for seniors less comfortable with typing.

👥
Multi-User Account Management

Role-based accounts for Admins and Users — supporting seniors, caregivers, and healthcare professionals from a single deployment with individual data privacy.

Persistent Memory Across Sessions

The AI remembers fitness goals, workout history, and user preferences across separate chat sessions — so every conversation picks up where the last one left off.

🔌
REST API & Data Integration

A full REST API backend with endpoints to log, retrieve, update, and delete activities. Integrates directly with the AI interface and supports Apple Health data imports.

See It in Action

The Senior Health Assistant uses a conversational AI interface. Users type or speak to log activities, ask for progress summaries, and receive guidance tailored to their fitness level and history.

SC
Senior Fitness Coach
● Online — Powered by local AI
Good morning! Based on your last 7 sessions, your average walk is 28 minutes. Ready to set this week’s goal?
Yes — I want to reach 35 minutes by Friday.
Great goal! I’ll track your daily progress. Try adding 2–3 minutes each day. Shall I log today’s walk when you’re done?
Yes please. I just finished 30 minutes.
Logged ✓ — 30 min walk today. You’re 5 minutes away from your goal. Keep it up!
28mAvg. Walk
12Sessions
98%Uptime

Quick Facts

Who It Serves

  • Elderly individuals maintaining or improving fitness
  • Caregivers monitoring a loved one’s progress
  • Healthcare professionals recommending wellness tools

Accessibility

  • Simple, uncluttered chat interface
  • Voice input and spoken responses
  • Mobile and desktop accessible
  • Minimal typing required

Technology Stack

The platform is built entirely on open-source components and runs on a self-hosted VPS — meaning all health data stays on your own infrastructure. No third-party cloud AI services required.

FastAPI Backend Ollama LLM Engine Open WebUI Gemma 3 4B Model SQLite Database Docker Ubuntu VPS Pipecat Voice Pipeline Pinecone Vector DB Python 3.10+ REST API (Bearer Auth) Adaptive Memory v3

The system architecture supports multiple concurrent users through a shared LLM API server, with session context managed per user in the database — no separate model instance needed per user.

Architecture Layers

  • Frontend: Open WebUI (web-based chat)
  • Backend: FastAPI REST application
  • AI Engine: Ollama running Gemma 3 4B
  • Database: SQLite (extendable to PostgreSQL)
  • Voice: Pipecat + Daily.co real-time audio
  • Memory: Adaptive Memory across chat sessions
  • Deployment: Docker on Ubuntu VPS (4-core / 16 GB)

Copyright LifeTech 2025-2030