Allrun Cloud Technology (Shenzhen) Co., Ltd.
Build a companion older adults can hear, understand, and choose to use
AI wellness companion network

Build a companion older adults can hear, understand, and choose to use

PTALK combines accessible voice interaction, dialect evaluation, familiar content, reminders, device management, and family or community workflows—with a clear boundary from medical care.

Large models, voice cloning, dialect recognition, and low-power endurance work together as one all-day companion system.
PTALK
AI Wellness
Care / Voice / Companion
Interaction
Voice-first
Language
Project-tested
Boundary
Non-medical
Care Capability

Build a more reliable wellness companion experience around senior needs

Age-friendly design, continuous companionship, and daily-life coverage turn AI wellness into something seniors can actually use and trust.

01

Designed for Seniors

PTALK AI: Designed Around the Real-Life Scenarios and Emotional Well-Being of Seniors
02

24/7 Caring Companion

Dedicated to providing round-the-clock caring companionship, ensuring seniors feel valued and supported in every moment of life
03

Age-friendly Approach

Integrated AI technology with age-friendly design, ensuring full functionality while remaining easy to use
04

Full Coverage Over Daily Scenarios

From practical daily help to emotional comfort, we cover every life scenario for senior friends
Technology Stack

Large models, voice cloning, and recognition technologies support the companion experience together

Language understanding, familiar voices, dialect recognition, and low-power design are organized as one technical support layer.

Overview
01
Use language understanding where it helps
02
Evaluate familiar voices responsibly
03
Test dialects with real speakers
04
Balance battery, volume, and portability
Project progress panel
Prototype
Prototype validation
Integration
Platform integration
Use language understanding where it helps
Evaluate familiar voices responsibly
Test dialects with real speakers
Balance battery, volume, and portability
01
Use language understanding where it helps
Organize daily questions, reminders, and content around the user's habits and the service boundary.
02
Evaluate familiar voices responsibly
Confirm authorization, disclosure, consent, and an appropriate use case before using a cloned or familiar voice.
Use language understanding where it helps
Powered by Leading LLMsPowered by leading LLMs including DeepSeek, PTALK delivers strong conversational AI skills
Fast and Accurate ResponsesAdvanced models deliver rapid, accurate responses for seamless interaction
Scenario-based Voice CommandsVoice commands for 100+ scenarios—from weather to music—get instant answers
Evaluate familiar voices responsibly
Voiceprint ExtractionAdvanced GAN spectral modeling extracts voiceprint features from speech samples
Model TrainingVoice recreation requires explicit authorization and project-specific evaluation of sample quality, model behavior, and target-device playback
High-Fidelity RestorationVoice similarity and naturalness must be evaluated with authorized samples and the target device; results should not be presented as identical to a real person
Test dialects with real speakers
Fusion Algorithm Integrated dialect models and semantic analysis deliver reliable dialect identification
Optimized RecognitionSenior-adapted voice logic solves "unclear speech, poor comprehension" problems
High AccuracyOver 95% accurate dialect recognition lets seniors converse in local language
Balance battery, volume, and portability
Battery ConfigurationPTALK AI companion features a 2000mAh large-capacity battery for extended usage
Efficiency ManagementFeatures a smart power management system that effectively reduces energy consumption and enhances battery efficiency
All-Day PowerCombines high-capacity battery and efficient system for 24-hour continuous operation
Application Scope

From care institutions to public welfare gifting, the service can enter multiple wellness settings

The same wellness device can extend into care communities, operator partnerships, gifts, and senior education scenarios.

01
Senior Care Facilities / Wellness Communities
Providing warm and intelligent senior care, PTALK voice cloning soothes, group chat connects and family apps monitor.
02
Channel Partnerships (Telecom Operators)
Partnering with China Mobile, China Telecom, and China Unicom, PTALK senior-focused AI companion with proven tech, simple service,...
03
Gifting & Public Welfare
As corporate gifts for employees parents or elderly-friendly supplies for public welfare organizations, PTALK AI companion devices...
04
Senior Universities & Training Institutions
Checking schedules, reviewing lectures, and access easily to classic courses like calligraphy in their own dialects, PTALK AI enha...
Age-friendly device selection

Separate home, community, and care-organization needs first

Selection should focus on whether older users can hear, read, and complete tasks with few steps, and how caregivers manage devices and content. Companion and reminder functions do not replace diagnosis, treatment, or emergency services.

Related sourcing terms
AI companion for older adultscare home smart devicescommunity care AI devicesage-friendly voice devicedialect voice companionsenior-friendly smart terminal
01

How do home and care-organization AI companion devices differ?

Home use emphasizes natural voice, simple reminders, family content, and consent. Organizations also need device groups, accounts and permissions, consistent content, network and charging operations, staff workflows, and incident handling. Map the actual service flow before comparing feature counts.

Final scope, pricing, and delivery timing depend on the brief, prototype evidence, and an agreed project boundary.
02

What determines the cost of an AI companion device for older adults?

Device form, display and controls, dialect and voice scope, connectivity, reminders and content, device management, privacy permissions, quantities, deployment training, and ongoing service all affect the estimate. Medical data, health recommendations, or emergency integration require separate compliance and responsibility review.

Final scope, pricing, and delivery timing depend on the brief, prototype evidence, and an agreed project boundary.
03

How should dialect support, large text, and low-step interaction be tested?

Invite target-age and dialect users to complete tasks under realistic noise, distance, lighting, and network conditions. Record wake-up, recognition, playback, text size, controls, navigation, and recovery from failure. Use task completion and recovery behavior as acceptance evidence, not only an office demo.

Final scope, pricing, and delivery timing depend on the brief, prototype evidence, and an agreed project boundary.