🌍 Earth's Pulse — Real-Time Global Sentiment Visualizer

HackTrent 2025 Submission
A real-time AI-powered platform that visualizes the emotional pulse of our planet using Reddit sentiment analysis.

---

📚 Table of Contents

• 🎯 Project Vision
• 🏆 HackTrent 2025 Prize Categories
• 🎯 The Problem
• 🌟 Key Features
• 🛠️ Technology Stack
• 💡 Our Solution
• 🚀 Quick Start Guide
• 📡 API Reference
• 🎨 How It Works
• 🌍 Geographic Coverage
• 🏆 HackTrent 2025 Alignment
• 🚧 Challenges & Solutions
• 🔮 Future Enhancements
• 👥 Team
• 📚 Documentation
• 📝 License
• 🙏 Acknowledgments
• 📞 Contact

---

🎯 Project Vision

Earth's Pulse transforms social media into a living, breathing emotional map of our planet.

By combining:

• real-time Reddit data
• AI sentiment analysis
• 3D globe visualization

…we give users a way to explore how communities across the world are feeling — right now.

---

🏆 HackTrent 2025 Prize Categories

This project competes for:

• ✨ Best Use of AI (Reach Capital)
  Advanced sentiment analysis using Hugging Face transformers and OpenRouter LLMs.

• 🎤 Best Use of ElevenLabs
  Natural AI-powered voice narration of city sentiment summaries.

• 🚀 General Category
  Full-stack innovation with real-world impact.

• 🤖 Best Use of OpenRouter AI
  Multi-model LLM integration for smart, contextual narratives.

---

🎯 The Problem

We live in a hyper-connected world, but we still struggle to answer simple questions like:

• How is the world feeling right now?
• Which cities are hopeful, anxious, or joyful?
• How do global events impact people emotionally across regions?

Existing tools focus on raw numbers (likes, shares, counts), not emotional climate.
Earth’s Pulse changes that by turning social media sentiment into an interactive, real-time emotional map.

---

🌟 Key Features

🔍 Real-Time Sentiment Analysis

• Live Reddit Data — Posts from 200+ cities worldwide via Reddit API
• AI-Powered Sentiment — Cardiff NLP RoBERTa model for robust classification
• Emotion Labels — Joyful (positive), Anxious (negative), Neutral

🌐 Interactive 3D Globe

• WebGL / Three.js Globe with realistic Earth textures and borders
• Color-Coded Mood Points –
  • 🟢 Positive
  • 🟡 Neutral
  • 🔴 Negative
• City-Level Detail for 200+ major cities across 6 continents
• Smooth Animations with clickable cities and zoom

🤖 AI-Generated City Summaries

• OpenRouter LLMs (Qwen, Gemma, Mistral, LLaMA models)
• Human-Readable Narratives that describe the emotional climate of each city
• Smart Caching (45-second TTL) to balance freshness and performance

🎙️ Text-to-Speech Narration

• ElevenLabs Integration for natural, human-like voices
• Multiple Output Modes: Base64, streaming audio, or URL
• Optimized TTS Settings (stability, clarity, style, speaker boost)

📊 Data Intelligence Layer

• MongoDB + Motor async storage for mood data points
• Background Refresh every 5 minutes (configurable)
• Historical Tracking via timestamps for trend analysis

🎨 Modern User Experience

• Next.js + TailwindCSS mobile-first UI
• Dark Theme with glassmorphism and gradients
• Interactive Popups & Live Stats
• Keyboard & Screen-Reader Friendly

---

🛠️ Technology Stack

Backend (FastAPI + AI/ML)

• FastAPI – Async Python web framework
• PRAW – Reddit API client
• Hugging Face Transformers – cardiffnlp/twitter-roberta-base-sentiment-latest
• Motor + MongoDB – Async document store
• OpenRouter API – LLM-powered summaries
• ElevenLabs API – Text-to-Speech
• httpx – Async HTTP client

Frontend (Next.js + 3D Globe)

• Next.js 14 – App Router, SSR/SSG
• React 18 + TypeScript 5
• TailwindCSS 3 – Utility-first styling
• Globe.gl + Three.js – 3D globe visualization
• Framer Motion – Animations
• Axios – API communication

DevOps & Infra

• Docker – Multi-stage container builds
• Docker Compose – Backend + Frontend + MongoDB orchestration
• Vercel – Frontend hosting
• Environment-based Config – secure API key loading

---

💡 Our Solution

Earth's Pulse is a full-stack AI platform that:

1. Collects
   Real-time Reddit data across 200+ cities.

2. Analyzes
   Each post with Hugging Face Transformers for sentiment.

3. Visualizes
   Global emotions on an interactive 3D globe.

4. Summarizes
   City-level emotional climates using OpenRouter LLMs.

5. Narrates
   Summaries using ElevenLabs Text-to-Speech for an immersive audio experience.

---

🚀 Quick Start Guide

✅ Prerequisites

• Python 3.9+
• Node.js 18+
• MongoDB (local or Atlas)
• API keys for:
  • Reddit
  • OpenRouter
  • ElevenLabs

---

1️⃣ Clone the Repository

git clone https://github.com/dikshithreddym/Earth-s-Pulse.git
cd Earth-s-Pulse

2️⃣ Backend Setup

cd backend---



# Install Python dependencies

pip install -r requirements.txt

## 🛠️ Tech Stack## 🏆 HackTrent 2025 Prize Categories```

# Create environment file

cp env.example .env



# Edit .env with your API keys:### Frontendearth-pulse/

# nano .env  (Linux/Mac)

# notepad .env  (Windows)

• Next.js 14 – React framework with server-side renderingThis project qualifies for:├── frontend/ # Next.js React application

Required Environment Variables:

# MongoDB

MONGODB_URL=mongodb://localhost:27017- **globe.gl** – 3D Earth visualization (Three.js based)├── backend/           # FastAPI Python application

DATABASE_NAME=earth_pulse

- **Tailwind CSS** – Modern, responsive styling

# Reddit API (Required for live data)

REDDIT_CLIENT_ID=your_reddit_client_id- **Framer Motion** – Smooth animations### 🥇 General Category├── docker-compose.yml # Docker orchestration

REDDIT_CLIENT_SECRET=your_reddit_secret

REDDIT_USER_AGENT=EarthPulse/1.0- **Plotly.js** – Interactive sentiment charts



# OpenRouter API (Required for AI summaries)└── README.md

OPENROUTER_API_KEY=your_openrouter_api_key

OPENROUTER_MODEL=qwen/qwen-2-7b-instruct:free### Backend



# ElevenLabs API (Required for voice synthesis)Full-stack application with innovative 3D visualization and real-time data processing```

ELEVENLABS_API_KEY=your_elevenlabs_api_key

ELEVENLABS_VOICE_ID=21m00Tcm4TlvDq8ikWAM- **FastAPI** – High-performance Python API



# Optional Settings- **PRAW** – Reddit API integration

ENABLE_BACKGROUND_REFRESH=true

REFRESH_INTERVAL_MINUTES=5- **Transformers (Hugging Face)** – Sentiment analysis (`cardiffnlp/twitter-roberta-base-sentiment-latest`)

ENVIRONMENT=development

```- **OpenRouter API** – LLM-powered summaries (LLaMA 3.1 8B Instruct)### 🤖 Best Use of AI powered by Reach Capital## 🛠️ Setup Instructions



**Start Backend Server:**- **ElevenLabs API** – Text-to-Speech narration

```bash

uvicorn main:app --reload --host 0.0.0.0 --port 8000- **MongoDB** – Data persistence with Motor (async driver)

• httpx – Async HTTP client

API will be available at: http://localhost:8000

Interactive docs at: http://localhost:8000/docs- Transforms how we understand global emotional well-being### Prerequisites

3️⃣ Frontend Setup### DevOps & Deployment

cd frontend

- **Docker** – Containerization for both frontend and backend

# Install Node dependencies

npm install- **Vercel** – Frontend deployment- Impacts learning (sentiment analysis education) and health (emotional climate awareness)- Node.js 18+ and npm/yarn



# Create environment file- **Railway/Render** – Backend deployment options

echo "NEXT_PUBLIC_BACKEND_URL=http://localhost:8000" > .env.local

- Python 3.9+

# Start development server

npm run dev---

🎵 Best Use of ElevenLabs- Docker (optional, for containerized deployment)

App will be available at: http://localhost:3000

🚀 Quick Start

Create .env.local file with:

GET /api/city/summary

Generate AI summary for specific city# NEXT_PUBLIC_BACKEND_URL=http://localhost:8000- MongoDB – Data persistence with Motor (async driver)cp .env.example .env

Query Parameters:

  - city: string           // City name (required)

  - limit?: number         // Max posts to analyze (default: 50)# Run frontend- **httpx** – Async HTTP client```



Response:npm run dev

{

  city: string,```

  summary: string,         // AI narrative about city mood

  statistics: {

    total_posts: number,

    positive: number,**Frontend runs at:** `http://localhost:3000`### DevOps & Deployment5. Fill in your API credentials in `.env`:

    neutral: number,

    negative: number,

    average_score: number```

  },### 4️⃣ Docker (Full Stack)

  sample_posts: Array<Post>,

  timestamp: string,

  data_source: 'reddit_api',

  ai_model: 'openrouter'```bash- **Docker** – Containerization for both frontend and backendREDDIT_CLIENT_ID=your_reddit_client_id

}

```docker-compose up --build



#### `GET /api/city/summary/audio````- **Vercel** – Frontend deploymentREDDIT_CLIENT_SECRET=your_reddit_secret

Convert city summary to speech

```typescript

Query Parameters:

  - city: string           // City name (required)---- **Railway/Render** – Backend deployment (production-ready)REDDIT_USER_AGENT=your_app_name/1.0

  - format?: 'base64' | 'url' | 'stream'

  - voice_id?: string      // ElevenLabs voice ID

  - model?: string         // TTS model override

## 🔑 API Keys Required- **GitHub Actions** – CI/CD pipeline (optional)TWITTER_BEARER_TOKEN=your_twitter_token

Response (base64):

{

  audio_base64: string,

  mime: 'audio/mpeg',Create a `backend/.env` file:MONGODB_URI=mongodb+srv://user:[email protected]/earthpulse

  summary: string,

  city: string,

  statistics: object,

  timestamp: string```env---OPENROUTER_API_KEY=your_openrouter_key

}

```# Reddit API (Required)



---REDDIT_CLIENT_ID=your_client_id```



## 🎨 How It WorksREDDIT_CLIENT_SECRET=your_client_secret



### 1. Data Collection PipelineREDDIT_USER_AGENT=EarthPulse/1.0## 🚀 Quick Start

```python

Reddit API → City-Specific Queries → 200+ Cities

├── Search: "{city}" (feeling OR mood OR today OR life...)

├── Filter: Last 24 hours, min 20 characters# OpenRouter API (Required for AI summaries)6. Run the backend:

├── Rate Limiting: Smart batching and caching

└── Output: Raw post text + metadataOPENROUTER_API_KEY=sk-or-v1-your-key-here

OPENROUTER_MODEL=meta-llama/llama-3.1-8b-instruct:free### Prerequisites```bash

2. Sentiment Analysis Engine

Cardiff NLP RoBERTa Model

├── Input: Cleaned post text# ElevenLabs API (Required for audio)uvicorn main:app --reload --host 0.0.0.0 --port 8000

├── Processing: Transformer neural network

├── Output: Label + Confidence ScoreELEVENLABS_API_KEY=sk_your-key-here

└── Mapping:

    ├── positive → joyful (0.3 to 1.0)ELEVENLABS_VOICE_ID=21m00Tcm4TlvDq8ikWAM- Node.js 18+```

    ├── negative → anxious (-1.0 to -0.3)

    └── neutral → neutral (-0.3 to 0.3)

MongoDB (Optional - uses in-memory if not provided)- Python 3.9+

3. AI Summary Generation

OpenRouter LLM Pipeline

├── Model: Qwen-2-7B-Instruct (primary)```- Docker (optional)Or with Docker:

├── Fallbacks: Gemma, Mistral, Llama

├── Prompt Engineering:

│   └── "Analyze {N} Reddit posts from {city}..."

│       "Write human-readable narrative..."### How to Get API Keys```bash

│       "Focus on lived experiences..."

├── Parameters:

│   ├── Temperature: 0.8 (creative)

│   ├── Max Tokens: 3001. **Reddit API** (Free)### 1️⃣ Clone the Repositorydocker build -t earth-pulse-backend .

│   └── Stop Sequences: [</s>, [/INST]]

└── Output: 4-6 sentence narrative   - Go to https://www.reddit.com/prefs/apps

• Create app → "script" typedocker run -p 8000:8000 --env-file .env earth-pulse-backend

4. Voice Synthesis

ElevenLabs TTS Pipeline

├── Model: eleven_multilingual_v2```bash```

├── Voice Settings:

│   ├── Stability: 0.62. **OpenRouter API** (Free $10 credit from MLH!)

│   ├── Similarity Boost: 0.7

│   ├── Style: 0.3   - Go to https://openrouter.ai/git clone https://github.com/dikshithreddym/Earth-s-Pulse.git

│   └── Speaker Boost: True

├── Input: AI-generated summary text   - Sign up and generate API key

└── Output: MP3 audio stream (Base64/URL/Stream)

  - Use MLH promo code for $10 creditcd Earth-s-Pulse### Frontend Setup

5. 3D Visualization

Globe.gl + Three.js Rendering3. **ElevenLabs API** (Free tier: 10k chars/month)```

├── Earth Texture: High-res day imagery

├── Country Borders: GeoJSON polygons   - Go to https://elevenlabs.io/

├── Mood Points:

│   ├── Color: Sentiment-based (Green/Yellow/Red)   - Sign up for free account1. Navigate to frontend directory:

│   ├── Size: Proportional to score strength

│   └── Animation: Smooth transitions   - Generate API key from Profile → API Keys

└── Interaction:

    ├── Click: Show city popup### 2️⃣ Backend Setup```bash

    ├── Hover: Highlight point

    └── Zoom: Auto-focus on selection4. **MongoDB** (Free)

• Go to https://www.mongodb.com/cloud/atlascd frontend

---

• Create free cluster

🌍 Geographic Coverage

• Get connection stringbash

200+ Cities Tracked Worldwide:

| Region | Cities | Examples |

| 🇺🇸 North America | 50 | New York, Los Angeles, Toronto, Mexico City |

| 🇧🇷 South America | 25 | São Paulo, Buenos Aires, Lima, Bogotá |

| 🇬🇧 Europe | 50 | London, Paris, Berlin, Madrid, Rome |

| 🇯🇵 Asia | 50 | Tokyo, Beijing, Mumbai, Seoul, Bangkok |## 📡 API Endpoints2. Install dependencies:

| 🇿🇦 Africa | 15 | Cairo, Lagos, Nairobi, Cape Town |

| 🇦🇺 Oceania | 10 | Sydney, Melbourne, Auckland, Brisbane |

Selection Criteria:### GET /api/moods# Create virtual environment```bash

• Population > 1M residents

• Active Reddit communities

• Geographic diversity

• Cultural significanceReturns real-time sentiment data pointspython -m venv venvnpm install

---

🏆 HackTrent 2025 Alignment```json# On Windows PowerShell:# or

Innovation & Creativity (25%)[

✅ Unique Concept - First real-time global emotional map using Reddit data

✅ 3D Visualization - Stunning WebGL globe with interactive mood points {venv\Scripts\Activate.ps1yarn install

✅ AI Integration - Multi-model LLM summaries + TTS narration

✅ Cross-Domain - Combines NLP, computer vision, and web technologies "lat": 43.6532,

Technical Complexity (25%) "lng": -79.3832,# On Linux/Mac:```

✅ Full-Stack - FastAPI backend + Next.js frontend + MongoDB

✅ AI/ML Pipeline - Hugging Face transformers + OpenRouter LLMs "label": "joy",

✅ Real-Time Data - Reddit API integration with background refresh

✅ 3D Graphics - Three.js + Globe.gl for complex visualizations "score": 0.85,source venv/bin/activate

✅ API Integrations - Reddit, OpenRouter, ElevenLabs, MongoDB Atlas

Functionality & UX (20%)

✅ Intuitive Interface - Click any city for instant sentiment details "source": "reddit",3. Run the development server:

✅ Responsive Design - Mobile-first approach with TailwindCSS

✅ Real-Time Updates - Live data refresh every 30 seconds "timestamp": "2025-11-09T12:34:56Z"

✅ Accessibility - Screen reader friendly, keyboard navigation

✅ Error Handling - Graceful fallbacks for API failures }# Install dependencies```bash

Impact & Relevance (20%)]

✅ Social Good - Mental health awareness through sentiment tracking

✅ Real-World Use - Disaster response, public health monitoring ```pip install -r requirements.txtnpm run dev

✅ Educational - Teaches data science, NLP, and visualization

✅ Scalable - Can expand to millions of users and cities

Presentation & Communication (10%)### GET /api/city/summary?city={city}&limit={limit}# or

✅ Clear Documentation - Comprehensive README with examples

✅ Demo Video - 3-minute walkthrough showcasing features

✅ Code Quality - TypeScript, type hints, clean architecture

✅ Live Demo - Deployed on Vercel for judgesAI-powered city sentiment analysis# Configure environment variablesyarn dev

Bonus: Sponsor Tech Integration

🏆 Google Gemini API - Alternative LLM for summaries (Gemma-7b-it)

🏆 ElevenLabs - Text-to-speech with voice synthesis jsoncp env.example .env

🏆 OpenRouter - Unified access to multiple AI models

🏆 Reach Capital AI - Sentiment analysis with Hugging Face{

🎯 Real-World Use Cases "summary": "Toronto's Reddit community shows a generally positive emotional climate...",

🏛️ Government & Policy Makers "statistics": {4. Open http://localhost:3000 in your browser

• Monitor public sentiment on policies and legislation

• Identify regions requiring social support programs "total_posts": 50,

• Track emotional impact of government decisions

• Emergency response during natural disasters "positive": 30,# Run backend

🧠 Mental Health Researchers "neutral": 15

• Study global emotional patterns and seasonal trends

• Identify cities with concerning sentiment trajectories "negative": 5,uvicorn main:app --reload### Docker Compose (Full Stack)

• Correlate emotions with world events and news

• Develop intervention strategies based on data "average_score": 0.42

📰 Journalists & Media

• Visualize public reaction to breaking news

• Create data-driven emotional climate reports "sample_posts": [...],

• Identify emerging social movements and trends

• Compare sentiment across different regions "data_source": "reddit_api",Run both frontend and backend together:

🎓 Educational Institutions "ai_model": "openrouter"

• Teach sentiment analysis and NLP concepts

• Demonstrate real-world AI applications}Backend runs at: http://localhost:8000

• Showcase data visualization best practices

• Research projects on social media analysis```

🏢 Social Good Organizations

• Target outreach to cities with negative sentiment

• Measure impact of community initiatives### GET /api/city/summary/audio?city={city}&format=base64

• Understand emotional well-being of communities

• Coordinate global mental health campaigns### 3️⃣ Frontend Setupdocker-compose up --build

🐳 Docker Deployment

Using Docker Compose (Recommended)

bash{bash

Build and start all services

• Services started: "mime": "audio/mpeg",cd frontend## 📡 API Endpoints

• MongoDB: localhost:27017

• Backend: localhost:8000 "summary": "...",

• Frontend: localhost:3000

📊 Performance & Statistics---# NEXT_PUBLIC_BACKEND_URL=http://localhost:8000 "lat": 40.7128,

• 200+ Cities tracked globally

• 50+ Posts analyzed per city per refresh

• < 1 second sentiment analysis per post## 🎨 How It Works "lng": -74.0060,

• < 2 seconds AI summary generation

• 5 minutes background refresh interval

• Real-time 3D globe visualization

• 30 seconds frontend auto-refresh### 1. Data Collection# Run frontend "label": "anxious",

Optimizations:

• Async I/O with FastAPI + Motor

• Connection pooling for MongoDB- Reddit API fetches posts from 200+ cities worldwidenpm run dev "score": -0.6,

• 45-second TTL cache for AI summaries

• Debounced resize handlers (150ms)- Posts are filtered by location, recency, and relevance

• Code splitting for globe.gl

• Smart city deduplication- Real-time updates every 30 seconds``` "source": "reddit",

🚧 Challenges & Solutions### 2. Sentiment Analysis "timestamp": "2024-01-15T10:30:00Z"

⚡ Challenge 1: Reddit API Rate Limits

Problem: Reddit limits requests to 60/minute

Solution: - Each post is analyzed using Hugging Face TransformersFrontend runs at: http://localhost:3000 }

• Implemented intelligent request batching

• City-level deduplication (one post per city)- Model: cardiffnlp/twitter-roberta-base-sentiment-latest

• Background refresh instead of real-time polling

• Smart caching with MongoDB- Returns emotion label and confidence score (-1 to 1)]

🗺️ Challenge 2: Accurate Location Mapping

Problem: Reddit posts don't include geolocation

Solution:### 3. AI Summary Generation### 4️⃣ Docker (Full Stack)```

• Curated list of 200 cities with verified coordinates

• City-specific search queries (e.g., "Toronto + mood")

• Manual validation of major urban centers

• Geographic diversity in city selection- OpenRouter LLaMA 3.1 model generates natural language summaries

⚡ Challenge 3: Real-Time Performance- City-specific prompts for contextual analysis

Problem: Heavy AI processing could block the UI

Solution:- Narrative style focusing on lived experiences```bash### GET /api/summary

• Async processing with FastAPI

• Background tasks for data refresh

• Client-side caching

• Progressive loading with skeleton screens### 4. Voice Narrationdocker-compose up --buildReturns global emotional summary:

🎨 Challenge 4: Three.js Compatibility

Problem: Globe.gl had missing module dependencies

Solution:- ElevenLabs TTS converts summaries to speech``````json

• Created custom patches for webgpu/tsl modules

• Implemented stub modules for missing dependencies- High-quality, natural-sounding voices

• Canvas size control for responsive rendering

• Multiple initialization attempts with timeouts- Auto-play with replay functionality{

🤖 Challenge 5: Consistent AI Summaries

Problem: LLM outputs can be unpredictable

Solution:### 5. Visualization--- "summary": "Today, South America is joyful while North America feels tense...",

• Detailed prompt engineering

• Multiple fallback models (Qwen → Gemma → Mistral)

• Stop sequences to prevent artifacts

• Post-processing to clean model tokens- globe.gl renders 3D Earth with Three.js "timestamp": "2024-01-15T10:30:00Z"

---- Color-coded sentiment points (green/yellow/red)

🔮 Future Enhancements- Interactive tooltips and city modals## 🔑 API Keys Required}

• Historical Trends - Track sentiment changes over time with charts- Real-time statistics dashboard

• Comparison Mode - Side-by-side city emotion comparison

• Twitter/X Integration - Combine Reddit + Twitter for richer data```

• Custom Date Ranges - User-selectable time periods

• Emotion Heatmap - Gradient intensity-based visualization---

• Multi-language Support - Analyze and translate non-English posts

• Real-time Notifications - Alert users for major sentiment shiftsCreate a backend/.env file:

• Mobile App - Native iOS/Android apps with push notifications

• User Accounts - Save favorite cities, custom alerts## 🎯 Use Cases

• Export Reports - Download PDF/CSV sentiment reports

🧑‍💻 Development Team

Developer: Dikshith Reddy M, Dev Patel Karan Majotra

University: Trent University

Event: HackTrent 2025 OPENROUTER_MODEL=meta-llama/llama-3.1-8b-instruct:free```

Timeline: Built in 24 hours

🎓 Education

---

🙏 Acknowledgments

• Teach sentiment analysis and NLP concepts

HackTrent 2025 Sponsors

• Reach Capital - For promoting AI innovation in education, health, and work- Demonstrate real-world AI applications# ElevenLabs API (Required for audio)### Frontend Tests

• ElevenLabs - For providing natural voice synthesis capabilities

• Google Gemini - For access to cutting-edge AI models- Showcase data visualization techniques

• OpenRouter - For unified LLM API access

• Major League Hacking (MLH) - For organizing this incredible hackathonELEVENLABS_API_KEY=sk_your-key-here```bash

Open Source Libraries### 🏢 Social Good Organizations

• Hugging Face - Cardiff NLP RoBERTa sentiment model

• Three.js Community - 3D graphics rendering engineELEVENLABS_VOICE_ID=21m00Tcm4TlvDq8ikWAMcd frontend

• Globe.gl - Interactive globe visualization library

• FastAPI - Modern Python web framework- Target outreach to cities with negative sentiment

• Next.js Team - React framework excellence

• Vercel - Seamless deployment platform- Measure impact of initiatives on public moodnpm test

Data Sources- Understand community emotional well-being

• Reddit - API access to community discussions

• Natural Earth - GeoJSON country boundary data# MongoDB (Optional - uses in-memory if not provided)```

• NASA Visible Earth - Earth texture imagery

Links

GitHub Repository: https://github.com/dikshithreddym/Earth-s-Pulse

API Documentation: http://localhost:8000/docs (when running locally)