01
Natural Language Search
Users can search with conversational phrases such as “movies like Interstellar” or “songs like On the Floor” instead of relying on rigid filters.
Cross-Media Recommendation Platform
Inspira
Inspira is a cross-media recommendation platform that helps users discover movies, television shows, books, and music through natural language searches.
Instead of searching separate platforms one at a time, users can explore recommendations across different media types, save favorites, and organize content into custom boards from one place. The platform also personalizes the experience using previous searches, saved content, and recent activity.
PHPMySQLJavaScriptBootstrapSQL
The Problem
Media discovery is fragmented across separate platforms.
People often move between different services when looking for a movie, television show, book, or song that matches a particular mood or reference. Traditional search tools also tend to depend on exact titles, genres, or keywords, which makes exploratory searches such as “movies like La La Land but happier” harder to express.
The Solution
One discovery experience across movies, shows, books, and music.
Inspira interprets natural-language searches and identifies the intended category, reference title, mood, and supporting keywords.
The platform then retrieves and ranks relevant results while keeping recommendations within the requested media category when needed.
Signed-in users can save favorites, create boards, revisit previous searches, and receive recommendations influenced by their activity.
Application Walkthrough
From a natural-language query to a personalized collection
02
Cross-Media Recommendations
Search results are presented as visual recommendation cards for movies, television shows, books, and music.
03
Save and Organize Recommendations
Users can add content to Favorites or save it directly to an existing board without leaving the recommendation page.
04
Custom Boards
Users can create personal collections and organize media by genre, mood, theme, or any category they choose.
05
Board Management
Each board can be renamed, updated, or deleted, and users can remove individual saved items from a collection.
Personalized Experience
Recommendations shaped by user activity
After signing in, users receive a personalized dashboard with recommendations based on favorites, recent searches, and previously viewed content. The dashboard also gives quick access to search, boards, and saved media.
Tools & Technologies
Frontend
PHP • JavaScript • HTML5 • CSS3 • Bootstrap
Database
MySQL • MariaDB • SQL
Application Features
Natural-language search • Personalized recommendations • Favorites • Custom boards
External Data
TMDB • Spotify • Google Books • NYT Books
Key Engineering Decisions
Built around personalization, category control, and reusable media data.
Search intent is separated from result retrieval so the application can identify the requested media type, reference title, mood, and keywords before ranking recommendations.
The database stores users, favorites, boards, board items, search history, and recent recommendation references so personalization can persist across sessions.
Recommendation behavior remains category-specific when a user asks for one media type, while still supporting cross-media discovery when the search is intentionally broad.
What I Learned
Building Inspira taught me how to connect a creative product idea to a complete database-backed user experience.
I worked through the full development process, including search behavior, data modeling, account features, personalization, API integration, and interface design. One of the main challenges was keeping results relevant across very different media types while still allowing users to search naturally. The project strengthened my experience with backend logic, relational databases, external data, and designing features that work together as one product.