What the app will — and won't — say.

MindfulGrove is an AI companion trained on therapeutic frameworks. This page explains how we shape the voice, what we refuse to do, and who reviews the work.

The voice

The companion draws on Cognitive Behavioral Therapy (CBT) and Emotion-Focused Therapy (EFT) frameworks, combined with a culturally aware foundation for the Indian context. It is trained to be warm, reflective, and non-prescriptive. It asks more questions than it answers.

What it will not do

• Diagnose a medical or psychiatric condition.
• Prescribe medication.
• Claim to replace professional therapy.
• Handle acute crisis situations — it will surface helplines instead.
• Pretend to be a human.

Cultural context

The app was built for India from the start. It understands family expectations, academic pressure, arranged-marriage dynamics, workplace hierarchies, and the social fabric in which these conversations actually happen — because a translation wrapper over an American therapy app cannot understand what “settling down” means in a Mumbai household.

Crisis handling

If the conversation contains language suggesting self-harm, suicidal ideation, or acute crisis, the app stops offering reflection and begins surfacing Indian helpline numbers. This is a hard-coded override, not a judgment call.

Clinical review

The therapy prompts, crisis-detection rules, and session structures are reviewed by licensed mental-health professionals before release and re-reviewed whenever substantial changes are made. Names and credentials will be listed here once reviewers consent to public attribution.

What we don't do with your data

Your conversations are not used to train AI models. Ever. This is a choice we made at the start of the project and it constrains the architecture — see the privacy policy for how it works under the hood.