Nidarsha

AI-Assisted Policy Insights and Recommendation System

Making government policies discoverable, understandable, and actionable.

---

Vision

Governments around the world invest billions into grants, subsidies, incentives, training programs, and support schemes intended to help citizens and businesses.

Yet a significant portion of these benefits remains underutilized—not because they lack value, but because the people who could benefit from them often never discover they exist.

Policies are scattered across numerous departments, written in complex administrative language, updated frequently, and rarely presented from the perspective of an individual or business trying to solve a real problem.

Nidarsha exists to bridge that gap.

Rather than asking users to search through government websites and policy documents, Nidarsha helps them understand:

• What policies are relevant to them
• Why those policies are relevant
• What benefits they may receive
• What requirements they must satisfy
• What actions they should take next

Our long-term goal is to transform government policies from static documents into practical decision-support tools.

---

Mission

Build an open, explainable, AI-assisted platform that helps individuals and businesses discover and utilize government policies more effectively.

The platform should reduce the effort required to:

• discover opportunities,
• understand eligibility,
• compare available options,
• prepare applications, and
• make informed decisions.

---

Initial Focus

The first release of Nidarsha focuses on Micro, Small and Medium Enterprises (MSMEs) in India.

The Indian MSME ecosystem contains hundreds of schemes across multiple ministries covering areas such as:

• Financial assistance
• Credit support
• Technology modernization
• Quality certification
• Export promotion
• Skill development
• Digital transformation
• Sustainability
• Innovation
• Manufacturing
• Entrepreneurship

Many businesses qualify for multiple schemes but remain unaware of them.

Nidarsha aims to improve that discovery process.

---

Project Status

This repository is currently in the planning and design phase.

The initial milestones include:

1. Understanding the policy landscape.
2. Building a structured policy knowledge base.
3. Designing an explainable recommendation engine.
4. Developing the first user experience.
5. Deploying an open-source MVP focused on Indian MSME policies.

---

Discovery Container

The discovery engine can run locally through:

python scripts/run_discovery.py

The main container pathway is Docker Compose:

docker compose run --rm discovery

For a local container smoke run that skips Google Sheets publishing:

docker compose run --rm -e NIDARSHA_PUBLISH_GOOGLE_SHEETS=false discovery

The Compose service mounts local artifacts/, logs/, and credentials/ directories into the container. Keep .env and Google service-account files outside git. Set NIDARSHA_PUBLISH_GOOGLE_SHEETS=false for smoke runs that should write only local manifest artifacts.