Research-Backed Mission

Evidence-Based AI Wellness for Everyone

We're not just building AI wellness tools – we're proving how supportive technology can be. Every feature is informed by peer-reviewed research, published studies, and real-world outcomes.

Participants felt significantly better
Users felt less worried
Published in NEJM AI

Our Research-Driven Story

VocalCalm was founded by researchers and clinicians who experienced the gap between mental health need and access. With over half of all people experiencing mental health challengesbut only 13 psychiatrists per 100,000 people globally, we knew traditional approaches couldn't scale.

Rather than building another AI chatbot, we committed to rigorous research from day one. The technology behind VocalCalm was part of the first-ever randomized controlled trial of generative AI wellness support, published in the New England Journal of Medicine AI, demonstrating measurable outcomes when humans supervise the experience.

The Science Behind VocalCalm

Our approach is grounded in three key psychological mechanisms identified in clinical research:

  • Radical Accessibility: 24/7 availability addresses the fact that 70% of mental health crises occur outside business hours
  • Anonymity Power: Removes stigma barriers, enabling honest self-disclosure crucial for therapeutic progress
  • Cognitive Mirror: Functions as a "bicycle for the mind," facilitating self-reflection and evidence-based skill building
51%
Felt Better
31%
Less Worried
106
Trial Participants
NEJM AI
Published In

Research Principles

Evidence-Based Development

Every feature grounded in peer-reviewed research and wellness science. Published in NEJM AI with measurable outcomes.

Safety Research

During the 2025 study, human supervisors intervened 15 times and corrected responses 13 times. Those learnings inform our automated signposting today.

Supportive Connection

Research shows users form meaningful supportive connections with AI comparable to traditional support sessions.

Research-Backed Results

Participants reported feeling significantly better and less worried in randomized controlled trials with 106 participants.

Clinical Evidence

Our platform is validated by peer-reviewed research across three critical dimensions: clinical efficacy, therapeutic alliance, and health equity.

Research-Backed Results

51%[1]

Participants reported feeling significantly better

g=0.61[2]

Moderate-to-large improvements in wellbeing for youth

d=0.82[7]

Large effect size at 8 weeks

Multiple randomized controlled trials demonstrate significant improvements. The first generative AI wellness support trial showed participants felt significantly better and less worried, with large effect sizes in engaged users.

Connection Quality

M=3.98 vs M=4.00[20]

Alliance score comparable to traditional support sessions

3-5 days[21]

Time to establish strong supportive connection

Trust and communicate with the system... to a degree that is comparable to working with a mental health professional[1]

Therabot Trial Participants, NEJM AI 2025

Research shows users form meaningful supportive connections with AI. Professionals were unable to distinguish AI transcripts from traditional support, rating AI quality as equal or higher.

Health Equity & Accessibility

235%[15]

Increase in referrals for non-binary individuals

31%[15]

Increase for ethnic minorities

67%[10]

More sessions attended vs traditional

The concept of it being anonymous really helped me open up and experience the stress-relieving nature of the app[14]

User Study Participant, 2024

AI wellness support dramatically improves access for marginalized communities. The NHS study found massive increases in care utilization among previously underserved populations, demonstrating anonymity as a powerful equity tool.

Clinical Evidence Sources

[1]Dartmouth Researchers (2025). First Therapy Chatbot Trial Yields Mental Health Benefits. NEJM AI.View source

Key finding: 51% average reduction in depression symptoms; 31% reduction in anxiety

[2]JMIR Research Team (2025). Effectiveness of AI-Driven Conversational Agents in Improving Mental Health Among Young People. Journal of Medical Internet Research.View source

Key finding: Moderate-to-large effect (g=0.61) on depression; large effect (g=0.74) in subclinical populations

[7]PATH Study Team (2024). A Randomised Controlled Trial of an AI-Enabled Mental Health Intervention. International Journal of Neuropsychopharmacology.View source

Key finding: Large effect size (d=0.82) in depression reduction at 8 weeks

[10]Eleos Health Research Team (2023). Effects of an Artificial Intelligence Platform for Behavioral Interventions. JMIR.View source

Key finding: 34% reduction in depression vs. 20% standard therapy; 67% more sessions attended

[14]NIH PMC (2024). User perceptions and experiences of an AI-driven conversational agent. NIH PMC.View source

Key finding: "The concept of it being anonymous really helped me open up and experience the stress-relieving nature"

[15]NHS Talking Therapies Research (2023). Closing the accessibility gap to mental health treatment with AI. medRxiv.View source

Key finding: 235% increase in referrals for non-binary individuals; 31% increase for ethnic minorities

[20]NIH PMC (2022). Evaluating the Therapeutic Alliance With a Free-Text CBT Chatbot. NIH PMC.View source

Key finding: Bond score with AI (M=3.98) statistically indistinguishable from human F2F therapy (M=4.00)

[21]Frontiers Digital Health (2025). Exploring user characteristics, motives, and expectations. Frontiers in Digital Health.View source

Key finding: Strong working alliance (M=3.76) established within 3-5 days

All research citations are from peer-reviewed journals and clinical trials conducted between 2021-2025.

Research Partnerships

Research Institute

Collaborative research partnership for ongoing studies and outcome validation

Licensed Supervisors

Board-certified psychologists who helped develop our research protocols and continue to advise on safety messaging

Academic Institutions

University partnerships for research validation and evidence-based development

Mental Health Organizations

Working with leading mental health advocates to improve access and reduce stigma

Research Advisory Board

Licensed clinicians and researchers helped design VocalCalm's prompts and safety guidance during the 2025 trial. Today they provide advisory input, but we do not run a live monitoring service—VocalCalm is a wellness app that signposts emergency resources rather than replacing professional care.

Join the Research

Be part of the future of mental healthcare. Complete the 30-question intake, explore a full week with your AI coach, and experience research-informed support that respects your privacy from day one.

*Results based on published NEJM AI research with 106 participants. Individual results may vary.

Not a substitute for professional medical advice, diagnosis, or treatment.