Muni: Consent as a Conversation
Leveraging AI to build conversation based consent sharing flows in 8 Indian languages for improved accessibility,understanding and reducing drop offs
Created on 4th November 2024
•
Muni: Consent as a Conversation
Leveraging AI to build conversation based consent sharing flows in 8 Indian languages for improved accessibility,understanding and reducing drop offs
The problem Muni: Consent as a Conversation solves
India's Account Aggregator Ecosystem revolutionized data sharing with 67.5M consents approved in the last one year. However this is only 55% of the total users.With almost half of the users dropping off we wanted to learn more about this, and find ways to fix this number.
After extensive market research and speaking to consumers we learnt that this gap stems from the following:
The current consent framework is not user friendly and involves a lot of technical jargon. For eg. FI type,Fetch Frequency etc, and for the common man this may come across as mambo jumbo resulting in them dropping off.
Also given the amount of digital fraud/scams it is very difficult for people to trust online platforms when it comes to their financial data.
-
Technical Barriers
- Assumes digital literacy
- Complex interfaces
- Limited support
-
Language & Accessibility
- Multiple languages make standard interfaces ineffective
- Traditional UIs exclude limited-literacy users
- Complex terminology creates barriers
-
Trust Issues
- Unclear understanding leads to hesitation
- Complex financial products breed uncertainty
- Impersonal digital consent
Our Solution: Conversational Consent
We believe consent goes beyond just yes/no—it's a dialogue. Our solution transforms the AA consent flow into an interactive conversation. We want this to be like a conversation with a real human helping the user along the way.
Features & Benefits
1.Builds Trust
- Natural conversations
- Native language
- Instant answers
2.Improves Access
- Voice-first
- Multilingual
- No tech expertise needed
- Familiar platforms (Whatsapp/Telephony)
3.Enhances Understanding
- Step-by-step guidance
- Context-aware help
Impact
Industry data shows 25-35% better conversion through:
- Natural interactions
- Quick responses
- Simple UX
- Personal guidance
We're democratising data sharing by making AA accessible to every Indian, turning consent into conversation.
Challenges we ran into
-
Selecting LLMs to Simplify AA Complexity
Our research identified user confusion with the Account Aggregator (AA) framework due to technical jargon and consent complexities. A conversational AI, using an LLM, could transform this process into a clear dialogue, which became our focus, with multilingual and contextual support prioritized. -
Opting for Sarvam AI as Our AI Partner
Sarvam AI stood out for India’s needs, offering multilingual support and integration with WhatsApp and telephony. While onboarding required extensive coordination, Sarvam’s adaptability allowed us to create a tailored consent experience, reaching users in familiar languages and conversational formats. -
Redesigning for an Accessible User Experience
To make AA consent accessible, we transformed it into a guided, voice-first dialogue. By simplifying jargon and using step-by-step guidance, our design clarified the AA process for all users, making complex interactions feel personal and straightforward. -
Managing Sarvam AI’s Platform Constraints
Though Sarvam AI was powerful, limited APIs and inbound-only communication required adjustments. We worked with Sarvam to develop custom APIs and redesigned our flow around these limitations, ultimately delivering a consent experience that bridged digital divides in India.
Tracks Applied (2)
Democratizing Insurance and Investment Advisory for Indians
Empowering Customers to Leverage their Digital Footprints
Technologies used