In May 2024, just like lakhs of students across India, I too finished my JEE examinations and was all set to go to college.
We embarked on the Adi Kailash Trip the very next day and returned 2 weeks later when I was faced with a crucial decision. Which college to choose? Which program to pursue?
There were more than 120 colleges, all of which were filled through the JoSAA counselling process, each offering approximately a dozen different streams to pursue. Which college and which stream to go for? Furthermore, there were a limited number of colleges I could get into, based on my rank, home state & category.
Thanks to the timely involvement and participation from both my parents, the three of us scoured platforms like CollegeDekho, Shiksha & CollegeDunia to find out more about each college that we found ‘interesting’. We also had to know the closing rank of programs in that particular college last year (the highest rank to get into that program in that college) and pragmatically aim for colleges that I had a chance of getting. Several YouTube videos of ‘campus tours, facilities, mess tours’ also had to be watched to get a sense of what we were getting into. This was easier said than done, with each platform offering scanty pieces of information, all of which were unverifiable. It was a confusing time with hundreds of different options and no ‘right answer’.
Going through this grind, I came to understand one thing. It’s just a game of information. The more information you have, the more informed decision you can make.
The Technical Challenge
Building a practical JoSAA guide isn’t just about displaying data. It’s about handling:
- 120+ engineering colleges across 4 different categories (IITs, NITs, IIITs, GFTIs)
- 1000+ engineering programs with varying durations and degree types
- 5000+ rank cutoff data points spanning multiple years, quotas, and categories
- Real-time filtering that needs to work across all this data instantly
- Intelligent rank-based predictions that factor in college prestige and historical patterns
- Displaying data in a meaningful way to make decisions quickly
This was about the same time I was exploring Phoenix LiveView and working with relational databases, and this seemed like the perfect chance to build an app that would present said data in real time. Around the same time, I was experimenting with structured outputs from Claude, having it seed my database with non-routine data. Writing that post spawned a new idea in my mind to somehow fetch data from verifiable sources that could be presented in a meaningful way.
The Accidental Beginning
One thing led to another, and soon I was laying the foundations for this application. I gathered data from the official JoSAA website, which was a treasure trove of information. The site had all the colleges, their programs, and rank cutoffs for each program across different years. Then I scraped the NIRF rankings from their official site, which gave me college rankings for the last few years. Lastly, I used Claude Haiku 3.5 to recieve structured data about each college, its location, year of establishment, and website.
Features
Thanks to LiveTable, all URLs remember their state, allowing you to share links with parents, friends, and teachers. The app is designed to be fast, responsive, and intuitive, with no loading spinners or page refreshes.
Rank Filter
The rank filter allows you to enter your rank and see only the colleges in which you can get atleast 1 program. It allows you to see all ranks acros the app color coded to indiciate yuor chances, taking into account the shift in cutoffs each year.For the sake of generality, all calculations are performed with respect to the Gender Neutral, Other state. The option to customize this will be available in future iterations. Based on your rank, all ranks are color coded with Green- Perfect! Orange- Maybe, Red - Might be difficult.
Real-Time College Discovery
Browse all 120+ colleges with instant filtering by:
- College type: IIT/NIT/IIIT/GFTI with dedicated toggles
- NIRF ranking tiers: Top 10, Top 25, Top 50, etc.
- Location-based search: Find colleges in specific states & cities
- Name search: Instant text-based filtering
All with millisecond response times. No loading spinners. No page refreshes.
Smart Program Exploration
Explore all 1000+ engineering programs with:
- Duration filters: 4-year vs 5-year programs
- Degree type filters: B.Tech, Dual Degree, Integrated courses
- Rank-based eligibility: See only programs you can get into
- Cross-college availability: Find which colleges offer your preferred program
Side-by-Side Comparison
Compare up to 4 colleges with:
- Rank cutoffs for common programs
- NIRF rankings and trends
- Shareable URLs - send comparison links to parents, friends
- Real-time updates as you add/remove colleges
Persistent state
Ranks are stored in liveStorage, allowing you to set your rank once and have it persist across all pages, always offering personalised results for your rank.
Official Data, Real Performance
Everything sourced directly from official JoSAA, NIRF sites - no unreliable third-party data. The entire app responds in milliseconds despite handling complex queries across 5000+ rank cutoffs with multiple joins.
Conclusion
What started as frustration with terrible counselling platforms turned into a comprehensive showcase of Phoenix LiveView’s capabilities. This project pushed liveview in exactly the right directions, custom interfaces, real-time data handling, and complex state management. I learnt a lot of stuff about LiveView and Phoenix, reaffirming my belief in project based learning.
For my fellow students facing the JEE counselling maze, this app is a game-changer. It transforms chaos into clarity, helping you make informed decisions with confidence. For the developers, its an advanced demonstration of livetable, and the many new things you can learn when you start to build real-world applications.