Nyaya Sahayak: An AI Agent for Accessible Legal Aid in India

Inspiration

The idea for Nyaya Sahayak came from our own struggle. Four of us were living in a rented flat when we suddenly found ourselves in the middle of a dispute with our landlord. Without prior notice, we were kicked out. We didn’t know the tenant rights, where to file a complaint, or even which authority could help us. In the end, we lost a significant part of our advance deposit money because we had no access to proper legal guidance. That frustration and helplessness made us realize how difficult it is for ordinary citizens especially those unfamiliar with legal systems to navigate disputes. Nyaya Sahayak was born from that pain.

What it does

Nyaya Sahayak is an AI-powered legal aid assistant that helps citizens understand and act on basic legal problems. Users can narrate or type their issues in plain language whether in Hindi, Tamil, Bengali, or English and the AI agent:

Classifies the problem (e.g., tenancy conflict, consumer complaint, domestic abuse).
Provides relevant legal information in simple terms.
Auto-generates draft petitions, affidavits, or RTI applications.
Suggests the nearest Lok Adalat, legal aid office, or NGO for follow-up.
Ensures privacy and trust with secure, certificateless authentication.

How we built it

We used IBM’s Granite models for natural language understanding and translation across Indian languages. The Agent Development Kit (ADK) orchestrates the workflow: intake → classification → document drafting → guidance. Data preprocessing pipelines handle multilingual input, and privacy-preserving cryptographic techniques ensure secure handling of user data.

Challenges we ran into

Legal complexity: Simplifying Indian legal processes into AI-understandable flows without oversimplifying.
Multilingual nuance: Capturing regional dialects while keeping accuracy.
Trust barrier: Citizens hesitate to rely on AI for legal matters, so security and transparency had to be central.

Accomplishments that we're proud of

Building an end-to-end automated workflow from citizen query to ready-to-use legal document.
Designing a system that can work for ordinary people, not just tech-savvy users.
Creating something that could genuinely reduce dependency on costly middlemen and empower communities.

What we learned

We learned that AI can’t replace lawyers, but it can dramatically lower the barrier for legal literacy. We also realized how important user empathy is when designing for vulnerable groups. Technology alone doesn’t solve problems context, culture, and trust matter equally.

What's next for Nyaya Sahayak

Expanding to cover more legal domains such as labor rights and women’s safety.
Building voice-first interfaces for low-literacy users in rural areas.
Partnering with NGOs and state legal services authorities for real-world pilot deployments.
Exploring integration with India’s e-Courts system to take automation beyond document drafting.