NAVARASA AI

Results
Movie Cards for each movie
Actor/Actress filmography details

Inspiration

Movie discovery has become increasingly overwhelming in the era of countless streaming platforms. Users often spend more time deciding what to watch than actually enjoying a film. Existing recommendation systems primarily depend on watch history and fail to understand natural language, emotions, or multilingual preferences.

Inspired by the ancient Indian concept of Navarasa—the nine universal human emotions—we wanted to create an AI-powered cinema companion that understands what users truly feel like watching. Our vision was to combine artificial intelligence with cultural storytelling, making both regional and global cinema easily discoverable through intelligent, personalized recommendations.

What it does

Navarasa AI is an AI-powered multilingual cinema discovery platform that helps users discover movies, TV series, and anime using natural language.

Users can search with simple conversational queries such as:

"Scary Malayalam movies from the 90s"
"Top-rated Korean dramas"
"Feel-good Tamil comedy movies"

The platform intelligently understands user intent and delivers personalized recommendations complete with AI-generated explanations, posters, trailers, cast and crew information, streaming platform availability, trending content, actor search, wishlist management, and watch history—all within a fast and intuitive cinematic interface.

How we built it

We built Navarasa AI as a full-stack AI application using modern web technologies and intelligent workflow automation.

The frontend was developed using React, TypeScript, Tailwind CSS, and Vite to deliver a fast, responsive, and cinematic user experience.

The backend leverages n8n to orchestrate AI workflows and API integrations. We used Groq's LLaMA 3.3 70B model for natural language understanding and generating personalized recommendation explanations, while TMDb API provides comprehensive movie, TV series, cast, trailer, and streaming data.

Natural language queries are processed through an AI intent extraction layer, dynamically routed through multiple workflows, enriched with real-world movie metadata, and finally returned with personalized AI-generated insights.

Challenges we ran into

Developing an intelligent recommendation system involved several complex challenges.

Understanding highly diverse natural language queries while maintaining accuracy was one of the biggest hurdles. We also had to correctly separate anime from live-action content, improve actor search despite spelling variations, generate personalized recommendation reasons without mismatching results, and optimize workflow performance while coordinating multiple API calls.

Building scalable AI workflows entirely using n8n while ensuring low response times required continuous testing, debugging, and optimization throughout development.

Accomplishments that we're proud of

We're proud of building a fully functional AI-powered recommendation platform that goes beyond traditional movie search.

Navarasa AI successfully combines natural language understanding, multilingual cinema discovery, personalized AI-generated recommendations, and real-world streaming information into a seamless experience.

We're especially proud of promoting regional Indian cinema alongside international content, creating an intuitive cinematic interface, and demonstrating how visual workflow automation with n8n can power production-ready AI applications.

What we learned

Building Navarasa AI significantly enhanced our understanding of AI system design, workflow automation, and full-stack application development.

We learned how to engineer reliable prompts for large language models, integrate multiple APIs into a unified intelligent pipeline, optimize AI performance for real-time responses, and design recommendation systems focused on user experience rather than just algorithms.

Most importantly, we learned that combining technology with cultural understanding creates far more meaningful digital experiences.

What's next for NAVARASA AI

Our vision extends far beyond this hackathon.

Future enhancements include voice-powered search, emotion-aware recommendations, mobile applications, collaborative watchlists, AI models that continuously learn user preferences, multi-user profiles, director and writer recommendations, and deeper personalization through watch history.

In the long term, we aim to expand Navarasa AI beyond movies into books, music, and podcasts, while collaborating with OTT platforms and regional film communities to become the world's most intelligent multilingual entertainment discovery platform.