Puja Altas: An Open Map for Durga Puja Hopping

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 2: Side Hustle

What I Built

Pujo Atlas is a browser-based map and itinerary planner for finding Durga Puja pandals in Kolkata and across West Bengal. It brings together a Kolkata puja guide and a government puja directory, with a combined view for exploring both datasets.

Visitors can search by pandal, locality, place name, coordinates, or DIGIPIN; browse nearby results on a clustered map; and save pandals to a separate hopping list to plan a route. The goal is to make it easier for residents and visitors to discover pandals and plan a day of hopping without switching between spreadsheets, map searches, and handwritten lists.

The app is currently a free community-facing project. It has no paying users or revenue yet.

Demo

The Pujo Atlas Demo on GitHub Page:

codebysumit.github.io

Code

codebysumit / pujo-atlas

Pujo Atlas is a static, browser-based map and itinerary planner for Durga Puja pandals in Kolkata and West Bengal.

Pujo Atlas

Pujo Atlas is a static, browser-based map and itinerary planner for Durga Puja pandals in Kolkata and West Bengal

Run locally

The app fetches its JSON data, so serve this folder over HTTP instead of opening index.html directly. From the project folder, run:

python -m http.server 8000

Then open http://localhost:8000.

OpenStreetMap is the default map and works without an API key. The government directory is loaded on demand and is about 12 MB. Records outside the supported West Bengal map bounds are omitted from the map.

Datasets and privacy

The JSON files used by the app are in data/:

• pujo_atlas_plan_city.json contains the Kolkata Puja Guide records. Its zone labels use the Kolkata Puja prefix, such as Kolkata Puja North.
• pujo_atlas_plan_government.json contains the government directory. President names and contact numbers have been removed from the records.

The app does not show or search government-directory president…

View on GitHub

How I Built It

Pujo Atlas is a static web app built with HTML, CSS, and vanilla JavaScript. Leaflet renders the OpenStreetMap view, and Supercluster groups dense sets of pandal markers. The app loads local JSON datasets, filters and sorts results in the browser, and stores hopping lists in browser storage.

The location search accepts place names, latitude/longitude, and DIGIPIN codes. Place suggestions use Photon with OpenStreetMap data; coordinates and DIGIPIN can be resolved locally. A searched location can show nearby pandals within 15 km. An optional Google Maps provider is available when configured with a browser-restricted API key.

Open-source AI disclosure: This version does not use an open-weight model, an open-source AI agent framework, or local inference. Open-source AI is not currently at the core of this project, so this draft does not yet meet the challenge's central technical prompt. I would need to build and document a real AI feature before presenting Pujo Atlas as a qualifying entry; I am not claiming that AI was used here.

Why Does Open Innovation Matter?

This version demonstrates the value of open mapping tools and data: a small browser app can offer searchable maps and useful planning features without requiring a proprietary map platform for its default experience. The source and data formats also make the project easier for others to inspect, adapt, and extend.

Open-source AI has not made anything possible in this version because it has not been integrated yet. I would rather be clear about that gap than suggest that the existing search and mapping features are AI-powered.

Prize Categories

I’m entering Best Use of GitHub Copilot. I used Copilot during development to help build and refine Pujo Atlas. I did not use GitHub Actions to automate the project or Copilot to review contributor pull requests.