Challenge background
Accessing public services is still a complex and frustrating process in many countries. Citizens often need to navigate multiple systems, fill out repetitive forms, and rely on manual steps that slow down the entire process. In many cases, digital platforms exist but are not intuitive, not accessible to everyone, or not designed around real user needs. Bhutan is a strong example of this challenge. While the country is moving toward digital transformation, many public services still involve fragmented processes, limited system integration, and barriers for users who are not comfortable with traditional digital interfaces. This creates friction for citizens, especially in rural areas where accessibility and usability are critical. Instead of jumping directly into building a complete solution, this project focuses on something more valuable: exploration and validation. The goal is to understand whether voice-enabled interaction, conversational AI, and agent-based workflows can realistically simplify how citizens request public services.
The problem
Accessing public services is still unnecessarily complicated. Users often need to go through multiple steps, systems are not fully connected, and digital platforms are not always accessible to everyone. Before attempting to build a complete solution, it is necessary to validate whether voice-enabled interaction, conversational AI, and agent-based workflows can realistically reduce friction, simplify the user experience, and improve accessibility.
Goal of the project
The main goal of this project is to explore the feasibility of using conversational AI, voice interaction, and agent-based workflows to simplify how citizens access public services.
Main Objective
Explore whether voice-enabled conversational systems and agent-based workflows can realistically improve accessibility and reduce complexity in public service requests.
Specific Objectives
- Understand the main friction points in public service access
- Identify one or two realistic use cases to explore
- Test different technical approaches (voice, chat, agents, workflows)
- Explore how service requests can be classified and routed
- Build small prototypes to validate assumptions
- Define a clear foundation for Phase 2 experimentation
Project timeline
- 1
Week 1
Week 1:Understanding the Problem
- Focus on understanding how citizens currently access public services and identifying the main friction points.
- Participants will explore possible use cases and define one or two scenarios to work on.
- 2
Week 2
Week 2: Exploring Possible Solutions
- The goal is to explore different approaches such as conversational AI, voice interaction, and agent-based workflows.
- Participants will design simple conversation flows and define small experiments to test ideas.
- 3
Week 3
Week 3: Experiments and Prototypes
- Instead of building a full product, the focus will be on testing ideas.
- Participants will build small prototypes, test request classification, and experiment with voice or chat interaction.
- 4
Week 4
Week 4:Validation and Next Steps
- The final week focuses on evaluating what worked and what didn’t.
- The team will define the most promising approach and prepare a clear direction for future phases.
What you'll learn
Participants will learn how to:
- Work on a real-world AI problem
- Explore technical solutions before building a full product
- Design conversational AI experiments
- Test the feasibility of AI solutions
- Work in a collaborative project with many contributors
- Move from problem understanding to technical experimentation