JoSAAkha

Inspiration

Many students, especially from Tier 2 and Tier 3 backgrounds, struggle with turning available information into a clear decision. They check cutoffs, placements, and curriculum separately without a unified view. We wanted to make this process simpler, clearer, and more confidence-driven.

What it does

JoSAAkha takes a student’s rank and preferences and shows relevant college and branch options.
It presents cutoff trends over the past five years to help users understand realistic chances.
Users can explore placements, departments, curriculum details, supported by a multilingual chatbot.

How we built it

We collected and structured JoSAA cutoff and college data into a consistent format. A clean interface was built to allow filtering, comparison, and trend visualization. We integrated a multilingual chat system to explain curriculum and college insights clearly.

Challenges we ran into

We faced issues while serving the endpoint of quantised Sarvam model even after registering it.We had to present complex information without making the interface overwhelming. Supporting multiple languages while keeping responses accurate was also challenging.

Accomplishments that we're proud of

We created a platform that simplifies an important decision by bringing together rank-based insights, cutoff trends, and college details into one continuous experience, making comparisons more meaningful and intuitive while keeping the overall flow simple yet insightful.

What we learned

We learned that clear presentation matters as much as accurate data, and that users value systems which reduce confusion and guide decisions effectively, while also gaining hands-on experience with the Databricks platform and MLflow to manage data workflows and experiment tracking in a more structured way.

What's next for JoSAAkha

We plan to extend this approach to other exams and admission systems. We aim to improve personalization and prediction accuracy further. We will expand the multilingual experience to reach more students across India.

Built With

• databricks
• python
• rag