TraveLens

Inspiration

As a group of girls who share a deep love for travel, we found that traditional search engines often fail to capture the "feel" of a place. We are constantly pinning dreamy landscapes, chic cafe interiors, and stunning architecture, but the journey from a beautiful image to an actual plane ticket is often filled with guesswork. We wanted to create a tool for people like us—travelers who prioritize the aesthetic and atmosphere of a destination. We wanted a way to take a vibe-heavy mood board and instantly turn it into a list of real-world locations that match that exact aesthetic energy we're always looking for.

What it does

• TraveLens acts as a bridge between visual inspiration and logical planning:

• Image Analysis: Users can upload photos or link a Pinterest board.

• Aesthetic Decoding: The system uses vision-to-text AI to identify architectural styles, natural landscapes, and color palettes.

• Smart Matching: It ranks real-world destinations using a matching algorithm to see how well they align with the input images.

• Travel Logistics: For every recommendation, it provides essential details like the closest IATA airport code and a personalized pitch on why that city fits your style.

How we built it

We built TraveLens as a full-stack app with a React frontend and a FastAPI backend. The core is a two-step AI pipeline:

• A vision model analyzes images and extracts travel vibe and features
• A second model generates and ranks destination recommendations We also integrated Pinterest scraping to turn boards directly into AI input, and designed a clean UI for a smooth user experience.

Challenges we ran into

One of the biggest challenges was fine-tuning the prompt so the AI could consistently follow our “master prompt” and return structured, usable results. Many models produced outputs that were too vague or inconsistent, so we had to switch to a paid model to achieve the reliability and control we needed. We also faced challenges with:

• Integrating multiple systems (React frontend, FastAPI backend, AI APIs)
• Handling unstable data sources like Pinterest scraping
• Managing dependency conflicts and environment setup

Accomplishments that we're proud of

We built a complete pipeline that turns visual inspiration into real travel recommendations. We’re proud of combining AI, frontend, and backend into a working product and making something as subjective as “vibe” actually usable.

What we learned

We learned that working with AI requires strong prompt design and structured outputs. We also improved our skills in full-stack development and learned how to connect multiple systems into one cohesive product.

What's next for TraveLens

• User Authentication: Implementing secure login so users can sync their wishlists across multiple devices.

• Expanded Sources: Adding support for Instagram links and direct gallery syncing.

• Social Features: Allowing users to share their "Travel Vibe" reports and curated wishlists with friends.

Enhanced AI: Refining the recommendation engine to include real-time weather and seasonal suggestions.

Built With

• css3
• fastapi
• github
• html5
• javascript
• node.js
• npm
• python
• react
• scraping
• skyscanner
• unsplash
• vite