AI Impact Summit

Global AI Impact Commons

AI Agents and Assistant for Digital Banking, Government Services, and Airline Customer Support

Bhutan flag

Bhutan

Financial Inclusion

Public Sector

Moderate

Implementing Organisation

No Mind Bhutan

Bhutan, Thimphu, Thimphu

Private Sector

Implementing Point of Contact

Ugyen Dendup

Co-founder and CEO

Contributor of the Impact Story

Carnegie India

Year of implementation

2022

Problem statement

Banks, government institutions, and airlines increasingly rely on digital channels to deliver essential services, yet many continue to operate on legacy systems with limited customer support capacity. Customers, citizens, and passengers often face long wait times, unclear processes, and inconsistent information when accessing banking services, government programs, or airline support. In banking and government services, users struggle with digital onboarding, service eligibility, documentation requirements, policy interpretation, and application status tracking challenges that disproportionately affect rural and underserved populations. Airlines face high volumes of repetitive customer queries related to bookings, cancellations, refunds, baggage, and flight schedules, leading to operational strain and reduced service quality. This use case deploys AI-powered digital service assistants for banks and government institutions, alongside AI customer support agents for airlines. Built using Natural Language Processing, these AI agents provide instant, multilingual, conversational assistance, resolve high-frequency queries at scale, and escalate complex cases to human agents when required, while integrating with existing systems.

Impact story details

NoMindBhutan is a Bhutan-based AI startup focused on building practical, human-centered artificial intelligence solutions for banks, government institutions, airlines, and education providers. Founded by young technologists and entrepreneurs, the organization addresses real-world service delivery challenges in regions with limited digital infrastructure and legacy IT systems. The organization specializes in AI-powered conversational agents, digital service assistants, and analytics platforms that integrate seamlessly with existing systems. NoMindBhutan emphasizes responsible and cost-efficient AI adoption by optimizing payloads, reducing system complexity, and ensuring privacy-aware deployments suitable for developing-country contexts. Since 2022, NoMindBhutan has deployed AI agents for customer support, digital service guidance, and internal knowledge access across banking, public services, and aviation. These solutions leverage Large Language Model (LLM), Retrieval Augmented Generation (RAG) and Natural Language Processing (NLP) to enable multilingual, always-available assistance while generating actionable insights for service improvement. Beyond enterprise deployments, NoMindBhutan is committed to digital inclusion and equitable access to services. Its approach positions AI as an assistive layer that augments human teams, improves service quality, and scales access without replacing human decision-making.

AI Technology Used

Natural Language Processing
Machine Learning

Large Language Model, Retrieval Augmented Generation

Key Outcomes

Efficiency

Productivity, Access

Reach, Inclusion

Equity, User Experience

Satisfaction, Accuracy

Quality Improvement

Narrative Outcome

Public institutions, banks and airlines in Bhutan operate on legacy systems with limited customer support capacity. Citizens face long wait times, unclear processes, and inconsistent information. No Mind Bhutan has deployed AI agents across banking, government, and airline services, serving over 140,000 users and resolving over 725,000 queries. AI-handled customer support interactions have grown by over 400 percent. The platform demonstrates how AI can be used to improve the customer experience without additional staffing needs at scale.

Impact Metrics

AI agents deployed across banking, government, and airline services

Baseline Value

Most services were primarily human-led support systems Number of AI agents

Post-Implementation

7 AI agents were deployed across banking, government, and airline services

Internal Monitoring

Total users supported through AI digital service assistants

Baseline Value

Limited digital self-service access Number of users

Post-Implementation

140 ,280 users were served through AI digital service assistants

Internal Monitoring

Total banking, government, and airline queries resolved

Baseline Value

Queries were resolved manually with long resolution times Number of queries

Post-Implementation

725 ,496 queries were resolved across services

Internal Monitoring

Year-on-year growth in AI-assisted service usage

Baseline Value

Usage statistics from 2024 were considered Percentage

Post-Implementation

More than 38.9% growth was achieved, 39,300 additional users benefitted from the service Percentage

Internal Monitoring

Increase in depth of user engagement

Baseline Value

Engagement was 1.43 messages per user in 2024 Messages per user

Post-Implementation

In 2025, AI agents led to increased user engagement with 5.17 messages being sent per user Messages per user

Internal Monitoring

Growth in AI-handled customer support interactions

Baseline Value

Customer support interaction volume from 2024 Percentage

Post-Implementation

AI-handled customer support interactions increased by more than 402.4% (581,095 additional messages) Percentage

Internal Monitoring

Implementation Context

Deployed

Bhutan

Bank customers, airline passengers, and citizens accessing government services, including youth, rural users, and underserved populations

Key Partnerships

Banks, government agencies, airlines, and private-sector technology partners

Replicability & Adaptation

Moderate

1. Localization for language, regulatory context, and service workflows is recommended. 2. Integration requirements may vary depending on existing banking, government, or airline legacy systems

* The data presented is self-reported by the respective organisations. Readers should consult the original sources for further details.