Skip to content

Civic AI and retrieval-augmented generation · 2025 – present

Rakshak AI

Citizens in rural Nashik need a simpler way to reach police services, file complaints and get legal guidance — in Marathi or English.

My role
Co-developer. Frontend and application development.
Team
Built by a multidisciplinary team responsible for AI, backend, data and system architecture, in collaboration with Nashik Rural Police.
Status
Deployed; evolving toward voice / IVR access
Technologies
Retrieval-augmented generationLLM applicationsWhatsAppVoice / IVRReact

Key contribution

Contributed to the frontend and application layer of a bilingual conversational system that maps police procedures and citizen-service intents to guided workflows.

How it fits together

  1. Who

    • Citizen

  2. Channel

    • WhatsApp
    • Voice / IVR (evolving)

  3. Conversation

    • Frontend and application layer (my contribution)
    • Bilingual dialogue (Marathi, English)

  4. Knowledge

    • Retrieval over police procedures and legal information

  5. Services

    • Complaint guidance
    • Lost-item reporting
    • Legal guidance
    • Station discovery
    • Emergency assistance
My contributionConceptual diagram from the project description — not the deployed architecture.

What it does

  • Complaint guidance
  • Lost-item reporting
  • Legal guidance
  • Police-station discovery
  • Emergency assistance

Problem and context

Rakshak AI was co-developed with Nashik Rural Police to improve citizen access to police services. Many interactions with a police station — knowing where to go, what a complaint needs, what the law says — are information problems before they are enforcement problems.

Users and stakeholders

Citizens seeking police services in Marathi or English, and the Nashik Rural Police, whose procedures and service intents define what the system can do.

My responsibilities

I contributed to the system’s frontend and application development. AI, backend, data and system-architecture responsibilities sat with other members of the team; this page does not claim them.

Approach

Police procedures, citizen-service intents and legal information were mapped into a bilingual conversational system built on retrieval-augmented generation, so responses draw on retrieved procedural and legal material.

The service launched as a citizen-facing WhatsApp platform and has since evolved toward voice / IVR access, so that people who do not use text chat can still reach it.

Outcomes

The CV reports 5,000+ users, 51,000+ citizen conversations and 80+ complaints and reports processed or resolved. These are usage figures; they are not a measured estimate of public-safety impact.

The project received institutional and public-sector recognition, including presentation to the Governor of Maharashtra.

Limitations and open questions

Usage counts show reach, not answer quality. A fuller evaluation would measure answer correctness against police procedure, failure and escalation rates, and accessibility for users who rely on voice.

Reported usage

users
5,000+
First year after the 1 Sep 2025 launch, as reported by the team in its one-year update (Sep 2026).
police-related citizen conversations
51,376
Same one-year window. Conversations, not unique users.
complaints and reports received, assessed and resolved
86
Same one-year window, as reported by the team.

Rakshak AI is a public-sector deployment, so internal architecture, prompts and data are not published here.