Contact Us

How It Works

Research and Evidence:
AI Voice Biomarkers for Early Dementia Detection

EldrAIr is a pioneering voice-based AI platform designed to detect the earliest, subtlest signs of cognitive declineyears before typical clinical diagnosis. It doesn’t focus on what words you choose, but how you speak them: more frequent pauses, searching for terms, slowed pace. These voice biomarkers, imperceptible to the human ear, can be early indicators of issues like Alzheimer’s, mild cognitive impairment, or even mental health challenges.

Evidence is mounting in the scientific community that such subtle speech changes matter:

  • Speech-based AI has already shown promise: one study achieved over 78% accuracy in predicting progression from mild cognitive impairment to Alzheimer’s disease by analyzing speech patterns over six years
  • Researchers at UT Southwestern developed a system analyzing spontaneous speechevaluating features like speech motor control and idea densitythat performed well in detecting cognitive impairment, often through just a 1–2 minute recording
  • Extensive reviews of AI in Alzheimer's detection confirm that integrating acoustic and linguistic features frequently achieves accuracy above 85%, especially when using deep-learning or pre-trained language models
  • The broader trend is: AI can indeed listenand detecthealth conditions through voice, including early signs of Alzheimer’s, heart disease, and more

How EldrAIr Works

Browser based AI conversation – 5 Minutes of Natural Conversation
A brief, friendly, natural-sounding callno forced testing, no pressure. Just a casual chat that lasts about five minutes.

Voice Analysis – AI Listens for Voice Biomarkers
EldrAIr’s sophisticated Listens to the words, the delivery, but the delivery: pitch variation, pauses, hesitations, speech rate, filler words, etc. These features are known to correlate with cognitive changes long before they manifest in more obvious ways.

Detection Insights – Flagging Subtle Changes Early
The AI compares each call against baseline patterns and broader norms. Any deviationslike increasing pause lengths, reduced fluency, or slowing speechare flagged as possible early signs of cognitive decline or memory issues.

Notification – Early Alerts to Caregivers and Providers
Alerts are sent to family members, caregivers, or healthcare providersdepending on user preferences, all while preserving privacy and agency. This early notice allows for follow-up, meaningful assessments, and timely support planning.

01

Browser based AI conversation

5 minutes, natural
conversation.

02

Voice
Analysis

Listens to the words, the
delivery

03

Detection
Insights

Subtle changes flagged
before symptoms appear.

04

Notification

Carers and providers
receive early alerts.

Why This Matters

Timely, Scalable, and Equitable: A five-minute browser based AI conversation is accessible to virtually anyone, reducing barriers caused by geography, mobility, and affordability.

Proven, Voice-First Approach: AI analysis of speech for early dementia detection is already being validated in real-world studies and pilot projects.

Transformative Impact Potential: Detecting early-stage cognitive decline dramatically expands the window for interventionoffering families and clinicians more time to plan, support, and improve quality of life.