Digital Sahayak – AI Government Services Navigator

• 

  Digital_Sahayak_workflow

Inspiration

Many citizens struggle to access government benefits because they don’t know which schemes they are eligible for, or how to apply. Long processes, limited awareness, and language barriers make it difficult especially for rural and underserved communities. We wanted to solve this with AI by building a citizen-first assistant that makes schemes simple, accessible, and transparent.

What We Aim to Learn

-> How to use agentic AI for real-world automation workflows. -> Building multilingual AI assistants with IBM Granite models. -> Structuring complex government data into citizen-friendly formats.

How We Plan to Build It

-> Use IBM Granite LLMs for natural language interaction. -> Employ the Agent Development Kit (ADK) to automate tasks like eligibility checks, form filling, and appointment booking. -> Apply data preprocessing tools to prepare government scheme data. -> Design a simple web/mobile interface for citizens to interact with the assistant.

Challenges We Expect

-> Collecting and standardizing data across different government schemes. -> Ensuring recommendations are accurate and reliable. -> Designing a smooth user experience for citizens with limited digital literacy.

Built With

• adk)
• agent
• backend)
• development
• frontend)
• granite
• ibm
• kit
• llms
• python