Smart India Hackathon 2024 - Finalist

The Challenge

Participated in the Smart India Hackathon 2024 — a national-level innovation competition organized to solve real-world challenges.

Our team tackled a problem statement from the Delhi Government, focused on developing a predictive solution using machine learning models on diverse datasets.

My Role

I handled the complete fullstack development — integrating the machine learning models with a scalable backend and building the frontend interface for users.

We built an intelligent platform that provides accurate predictions based on dynamically changing datasets. The challenge was not just about building a model, but creating a production-ready system that could handle real-world data at scale.

The Team Dynamic

We were a team of 5 members, where I was the only 2nd-year student, collaborating alongside 3rd and 4th-year seniors. This was both challenging and incredibly valuable — I had to:

• Quickly adapt to the team's workflow and technical stack
• Take ownership of the entire web development pipeline
• Bridge the gap between ML models and user-facing application
• Collaborate effectively with seniors who had more experience

Being the youngest member pushed me to prove my capabilities and contribute meaningfully to the team's success.

Technical Implementation

Machine Learning

• Built predictive models using scikit-learn and TensorFlow
• Implemented data preprocessing pipelines for diverse datasets
• Model evaluation and optimization for production deployment

Backend

• FastAPI for high-performance API endpoints
• PostgreSQL for structured data storage
• Real-time data processing and model inference
• RESTful API design for frontend-backend communication

Frontend

• Next.js for server-side rendering and optimal performance
• React with TypeScript for type-safe component development
• Tailwind CSS for responsive, accessible UI
• Interactive dashboards with Chart.js for data visualization

What I Learned

This was a solid hands-on experience blending AI + web dev at scale. Key learnings:

1. Production ML is different — Model accuracy is just one metric; latency, scalability, and maintainability matter equally
2. Cross-functional collaboration — Working with data scientists and understanding their models to create seamless integrations
3. Time-boxed innovation — Building under pressure with tight deadlines teaches prioritization
4. Government tech challenges — Understanding how technology can solve civic problems at scale

The Impact

Reaching the finalist stage at a national-level competition validated our approach and hard work. More importantly, it showed that as a 2nd-year student, age is just a number — what matters is the willingness to learn, adapt, and deliver.

This experience solidified my interest in building AI-powered applications that solve real problems, not just tech demos.