MeldRX
Chikitsa
user registration 2
user registration
user registration 1
user registration 3
user login
Home page
our services
home page
Digital psychiatrist (consultant):Chatbot
voice assistant(emotion detection)
Open ended questions
Close ended questions
Wellness report (based on open and close ended questions)
image analyzer(medical report)
image analyzer(result)
image analysis (of an image with rashes on a hand)
Mood tracker
Meditation
user profile(wellness report)
user profile(meditation report)

Chikitsa: An AI-Powered Mental Health Revolution

Inspiration

The idea for Chikitsa was born during the Google Gen AI Exchange Hackathon (38,000 participants), where we—Vinayak Shinde and Vishal Misra, fourth-year engineering students at Mumbai University—saw a $1 billion challenge: mental health care in India is out of reach for millions. With just 0.75 psychiatrists per 100,000 people (National Mental Health Survey), soaring costs, and deep-rooted stigma, urban youth like our peers lack support. Earning 3rd place fueled our drive to harness predictive AI, blending tech with empathy to create a solution that could change lives.

What it does

Chikitsa is an AI-powered digital psychiatrist delivering affordable, private, and predictive mental health support, built to integrate seamlessly with healthcare systems via MeldRx app with EHR on SMART on FHIR standards.

what it offers:

AI Chat Support 🤖: Uses the Gemini API for real-time psychiatric consultations, predicting emotional states and providing empathetic responses 24/7.
Voice Interaction 🎤: Emotion-aware AI analyzes voice tone to classify moods (e.g., anxious, calm), tailoring predictive interventions.
Image Analysis 🏥: Processes medical reports or visual cues (e.g., stress indicators) to detect early mental health risks proactively.
Mood Tracking 📊: Daily check-ins analyze emotional patterns, predicting trends for personalized insights.
Guided Meditation 🧘: Suggests mindfulness exercises based on mood data, enhancing well-being.
Wellness Path 🚀: AI-driven questionnaires evaluate mental states, offering tailored feedback and plans.
Live Doctor Consultations 🩺: Video sessions with licensed psychiatrists connect AI predictions to human care.
Data Privacy & Security 🔒: End-to-end encryption safeguards data, compliant with FHIR R4 standards.
Corporate Wellness Solutions 🏢: Predictive analytics track and improve group mental health, like employee well-being. Chikitsa’s hybrid approach—AI predictions plus human trust—ranges from a free tier (₹0) to a full-featured version (₹1,000/month), targeting India’s urban youth.

How we built it

We crafted Chikitsa as a prototype during the Google Gen AI Hackathon, later adapting it for MeldRx to meet this challenge’s requirements:

Tech Stack:

Gemini API: Drives predictive AI chat and NLP, processing FHIR R4 data for consultations.
Deep Learning: Custom models classify voice emotions (e.g., anxiety) and analyze images (e.g., reports).
SMART on FHIR: Uses EHR Launch and CDS Hooks via MeldRx for healthcare system integration.
WebRTC: Powers secure, scalable video calls, hosted on AWS/Google Cloud.
FHIR R4 Workspaces: MeldRx’s backend manages patient data and analytics securely.

Process:

Conceived Chikitsa to address India’s mental health gap during the Google Hackathon.
Built the AI core with Gemini API, adding voice and image analysis via open-source tools.
Integrated SMART on FHIR with MeldRx, ensuring EHR Launch compatibility.
Created a React Native app for cross-platform use, tested with 50 beta users.
Vinayak focused on AI, Vishal on UI/UX and backend—collaborating as a student duo.

Challenges we ran into

Creating Chikitsa pushed our boundaries:

Emotion Detection Accuracy: Limited datasets made training predictive voice and image models tough within hackathon time limits.
FHIR Integration: Mastering SMART on FHIR and MeldRx’s FHIR R4 setup required a steep learning curve.
Video Call Reliability: WebRTC’s performance varied with bandwidth, complicating live sessions.
Privacy Compliance: Adding encryption and meeting FHIR security rules stretched our skills.
Time Pressure: Juggling studies and internships at Engagely AI with hackathon deadlines meant ruthless prioritization.

Accomplishments that we're proud of

Our hard work paid off in exciting ways:

Hackathon Win: Took 3rd place in the Google Gen AI Exchange Hackathon (38,000 participants), proving Chikitsa’s promise for a $1B issue.
MeitY Support: Selected by MeitY Startup Hub, India, gaining access to resources and recognition.
Google Nod: Invited to Google Delhi office, expanding our network and visibility.
User Feedback: 50 beta users loved it—all satisfied and wanting it available soon—showing real demand.
MeldRx Fit: Successfully launched Chikitsa via EHR Launch on MeldRx, a big leap for our team.

What we learned

This experience taught us a ton:

Predictive AI: AI can predict mental health risks, but needs lots of data and tweaking to shine.
FHIR Standards: SMART on FHIR makes healthcare apps connect seamlessly—a game-changer we now grasp.
User Design: Simple, caring interfaces beat complex ones—accessibility is key.
Team Power: Combining Vinayak’s AI know-how and Vishal’s coding skills made us unstoppable.
Grit: We thrived on limited resources, using creativity and open-source tools to push forward.