Building TrackNation: The Tech Stack, The Vision, and The Journey

Main Thread

Introduction: Why "Track" Nation?

Here is a problem that has plagued us for years: When a major issue unfolds in India—whether it's a policy change, an environmental crisis, or a high-profile legal case—where do you go to get the complete picture?

If you Google a topic like the aftermath of the Bhopal Gas Tragedy, the progress of the Ganga cleanup, or the ongoing air pollution crisis in Delhi, you will find scattered news articles, outdated Wikipedia entries, and opinion pieces. What you won't find is a single place that tracks the evolution of the issue over time.

TrackNation was born from this frustration. It is designed for subjects that evolve. It is a platform where articles serve as living records that update as new information surfaces. Whether it is tracking India's renewable energy targets, the status of major infrastructure projects like the Bullet Train, or the long-term impact of monsoon patterns on agriculture—we want one place where you can follow the complete story.

Today, I am excited to announce that TrackNation (tracknation.in) is officially live. I want to take you through the entire journey—from the initial idea to a production-grade platform.


The Vision: Two Pillars of TrackNation

Before diving into the technical details, let me explain what TrackNation is really about.

Pillar 1: Tracking Major Issues in India

The primary purpose of TrackNation is to serve as the definitive source for tracking ongoing issues in India. Think about topics like:

  • Climate & Environment: Air quality trends, river pollution levels, deforestation data.

  • Infrastructure Projects: Metro expansions, highway networks, smart city progress.

  • Social Issues: Education access, healthcare improvements, unemployment trends.

  • Major Cases: High-profile legal cases where developments span months or years.

For each topic, TrackNation provides a single, evolving thread. New information gets added, but the history is preserved. You no longer have to hunt through 50 separate news articles to piece together a timeline.

Pillar 2: Learning & Knowledge Retention

The secondary purpose is helping readers actually remember what they read. This is where our Active Recall Engine comes in—an experimental learning system based on spaced repetition (the same science that powers apps like Anki).

When you finish reading an article, the platform can quiz you on key facts. Days later, you get reminded to review. It is still being refined, but the goal is to transform passive reading into active learning.


The Architecture: Building for Stability

When I started TrackNation, it was a mess of spaghetti code running on whatever free tier I could find. Moving to production meant building a robust foundation. Here is the current setup:

The Tech Stack Visualization


Why This Stack?

Frontend: React + Vite on Cloudflare Pages

I have spent my corporate days working with various frontend frameworks, but for TrackNation, I wanted something modern and lean. React combined with Vite gives us:

  • Zero cold starts (unlike some "free" hosting services that sleep).

  • A global edge network, ensuring instant loads from anywhere.

  • Automatic HTTPS, preview deployments, and instant rollbacks.

  • Cost efficiency: Essentially free for our current scale.

Backend: FastAPI on EC2

While I have worked extensively with ExpressJS, for this project I wanted something type-safe and high-performance. FastAPI is perfect for building the complex connections required to link articles to ongoing updates:

  • Automatic OpenAPI documentation.

  • Pydantic data validation.

  • Async performance.

  • Hosting on EC2 gives us the control we need without the "sleeping" issues of serverless free tiers.

Database: MongoDB Atlas

Traditional SQL (rows and columns) can be rigid. Since our "Issue Tracking" style requires flexible data fields and different block types within an article, MongoDB's document-based style allows us to adapt the schema on the fly.

Why NoSQL? Try storing a rich-text article with embedded images, code blocks, variable sections, and nested comments in traditional SQL. You will quickly understand why we went with MongoDB.

Storage: Cloudflare R2

  • Cheaper than AWS S3 (no egress fees!).

  • Offers the same S3-compatible API.

  • Integrated CDN for global delivery.


The "AI" Factor: High-Speed Co-Development

Let’s address the elephant in the room: I used AI to write a significant portion of the code. But don't let that fool you—if you aren't sharp with system design and debugging, AI will lead you into a rabbit hole of broken logic.

AI is a powerful companion that speeds up the "grunt work," but I still had to do the heavy lifting when the AI's logic hit a wall. Understanding architecture, debugging edge cases, and making design decisions is still very much a human job. I will be writing a separate deep-dive soon on how to effectively manage AI to build a full-scale app.



The Features: What Makes TrackNation Different

1. The Block-Based Editor

Our editor isn't just a text box. It is a full block-based system (powered by TipTap/ProseMirror) that supports rich text, video embeds (YouTube), code snippets with syntax highlighting, and media management. Writers get a distraction-free environment that renders beautifully in both light and dark modes.

2. Active Recall & Spaced Repetition (Experimental)

This is our experimental learning feature. Authors can mark keywords, statistics, and key points within an article. After reading, users can practice recall. The system uses the SM-2 algorithm (the same one used by medical students) to schedule reviews at optimal intervals.

The Matching System is smart:

  • Fuzzy string matching catches typos (e.g., "photosythesis" still matches "photosynthesis").

  • Statistics match within tolerance ranges.

  • NLP weighting ensures you grasp the core concept, not just specific wording.

3. Aura Points: Gamification That Works

Learning is more fun when there is a scoreboard. Users earn "Aura Points" based on their recall accuracy.

Recall Accuracy

Base Points

90%+

50 points

75-90%

40 points

50-75%

30 points

30-50%

20 points

<30%

10 points

We also use Frequency Multipliers to reward consistent learners. Reviewing after 30 days yields 1.5x points. There is a maximum limit per session—no shortcuts, no cheating.

4. Playlists: Curated Learning Paths

Think of this as Spotify for articles. Users can create public or private playlists, organize stories in a specific order, and share them. This is perfect for courses, research collections, or "must-read" lists.

5. Role-Based Publishing Workflow

We built a complete content moderation system to ensure quality. Every story goes through a review process before going live.

  • User: Read, comment, vote, save.

  • Writer: Create stories, submit for review.

  • Moderator/Admin: Approve drafts, moderate comments, manage analytics.


The Technical Challenge: NLP Matching at Scale

Matching user recall responses against story content needs to be accurate but fast.

The Solution:

  • Client-side NLP processing to avoid server roundtrips.

  • Levenshtein distance for fuzzy matching.

  • POS (Part of Speech) tagging to weight important words.

  • Generic word filtering so that common words don't trigger false positives.


What's Next: The Roadmap

TrackNation is live, but I am just getting started. Here is what is on the horizon:

  • More Issue Coverage: Building out comprehensive threads on major ongoing issues.

  • Community Contributions: Allowing verified contributors to add updates to existing threads.

  • Mobile Experience: Optimizing for on-the-go reading and tracking.

  • AI-Powered Summaries: Auto-generating key points for quick catch-ups.

  • Notification System: Get notified when issues you follow have updates.


The Stack Summary:

Layer

Technology

Why

Frontend

React 18 + Vite + Tailwind

Speed, DX, flexibility

Backend

FastAPI + Python 3.11

Type-safe, async, autodocs

Auth

JWT + bcrypt + OAuth

Secure, scalable

Database

MongoDB Atlas

Flexible schema for all data

Storage

Cloudflare R2

Cheap, fast, S3-compatible

CDN

Cloudflare

Global edge, DDoS protection

Hosting

Cloudflare Pages + EC2

Best of both worlds

CI/CD

GitHub Actions

Automated testing and deploys



Join the Journey

TrackNation is still in its early stages—I am currently the primary user, testing and refining features daily. But the vision is clear: a platform where major issues in India don't get lost in the news cycle, and where learning becomes an active, not passive, experience.

Whether you are interested in following the evolution of national issues, or you want to try a new approach to retaining what you read, TrackNation is being built for that purpose.

Visit us at tracknation.in

Follow the development journey. Report bugs. Suggest features. The platform is live, and I'm building in public.

Tracking what matters.