Inspiration

Traditional travel planning suffers from several issues:

  1. Generic Recommendations — Most travel sites recommend the same popular destinations to everyone, regardless of personality or travel style
  2. Personality Mismatch — A bustling city hotel might be perfect for extroverts but exhausting for introverts. Standard filters can't capture this.
  3. Decision Fatigue — Too many choices without personalized guidance leads to analysis paralysis

What it does

Lotara is an AI-powered travel planning platform that creates truly personalized Vietnam itineraries based on your personality, not just your preferences.

How we built it

We built Lotara with a modern AI-first stack. A Next.js frontend captures users’ travel personality through a chat-based experience. A NestJS backend converts those answers into a multi-dimensional “vibe profile” and scores destinations using a compatibility algorithm. For deeper planning, a Python-based multi-agent AI system generates personalized, budget-aware itineraries. Everything is deployed on Vercel with full tracing and guardrails to keep outputs reliable and explainable.

Want to learn more about the idea and product direction? Take a look at our presentation.

https://docs.google.com/presentation/d/1Ansu167D5Knanb7HTQ4JJhk0wJEOFb-R7LyIBkLAOwQ/edit?usp=sharing

Interested in the technical architecture or how it was built? Explore the full implementation on GitHub.

https://github.com/nminhkhoi0818/lotara

Built With

Share this project:

Updates