Doot: India’s proposal for a national AI agent for every citizen

A sovereign AI architecture could reshape digital services—and potentially create a new distribution channel for travel

India’s Digi Doot whitepaper proposes a national AI architecture called Agent One, where every citizen would have a personal AI agent acting as their digital representative across online services. Instead of navigating apps, portals, and booking websites, citizens could interact with services through voice or natural language while the agent handles discovery, verification, and transactions. The architecture builds on India’s digital public infrastructure—including Aadhaar identity, DigiLocker documents, and the UPI payments network—and introduces a trusted agent layer designed to simplify complex digital interactions. For the travel industry, the concept signals how AI agents could become a new distribution interface connecting travelers directly with suppliers.

Key takeaways

• A personal AI agent for every traveler: The proposal envisions each citizen having a digital agent that can discover services, compare options, and complete transactions on their behalf, potentially becoming the primary interface for booking travel.
• Built on India’s digital infrastructure: The system integrates with Aadhaar for identity verification, DigiLocker for documents and agent memory, and UPI for payments, creating a trusted environment for AI-driven transactions.
• Dual-agent architecture enables privacy and scale: A phone-based agent runs locally on the user’s device to manage personal data and conversations, while a cloud-based replica performs complex reasoning and multi-service coordination.
• Travel booking demonstrated as a real use case: In the whitepaper example, an elderly traveler uses voice commands to plan a trip, discover transport and accommodation providers, verify eligibility for senior discounts, and complete bookings through secure digital payments.
• Hotels could connect directly to traveler agents: If adopted widely, citizen-owned AI agents could become a new distribution layer where hotel availability, pricing, and booking logic are accessed via APIs rather than traditional OTA search interfaces.
• Privacy-preserving personalization becomes possible: Verified credentials could allow travelers to prove eligibility for discounts—such as senior or accessibility rates—without revealing sensitive personal information.
• Human verification remains part of the booking flow: High-risk actions like payments require explicit traveler confirmation, maintaining consumer protection even in automated booking environments.
• Kumbh Mela 2027 planned as a large-scale test: The architecture could be tested during the world’s largest pilgrimage event, with goals including high task completion rates and significant reductions in coordination costs.
• A potential shift in travel distribution: If AI agents become widely adopted, the traditional relationship between travelers, OTAs, and supplier websites could evolve toward agent-to-service interactions.

Source: Doot via White Sky Hospitality (their weekly newsletter is highly recommended)

Enjoying this analysis? Hospitality.today delivers daily insights on hotel distribution, AI trends, and travel commerce — straight to your inbox. Subscribe for free at Hospitality.today →