Trek Sentry

Project Story: Trek Sentry

Inspiration

The inspiration for Trek Sentry came from observing the serious risks travelers face when critical information is scattered, delayed, or difficult to understand during natural disasters and extreme weather events. Incidents such as the Murree snowstorm demonstrated how the lack of clear, centralized travel alerts can lead to confusion and preventable harm. This highlighted the need for a unified system that delivers real-time, actionable, and easy-to-understand guidance. Trek Sentry was created to combine live global data, AI-driven reasoning, and multilingual accessibility, enabling travelers to make informed and safer decisions in rapidly changing conditions

What I Learned

Working on this project taught me how to integrate multiple technologies and handle real-time data. Key learnings include:

Fetching and processing asynchronous data from APIs using Python's aiohttp.
Integrating Google Gemini 3 for reasoning and generating clear insights from raw technical data.
Connecting TTS services to convert AI-generated summaries into audio alerts.
Handling front-end interactivity with HTML, CSS, and JavaScript.
Deploying on Vercel and managing cloud services efficiently.

How I Built It

The system pulls live data from multiple APIs weather, traffic, air quality, and earthquakes and passes it to Google Gemini 3 for analysis. Gemini generates situation summaries, preparation guidelines, and recommendations. Outputs can then be converted to audio and translated into 100+ languages. The front-end presents all this in a single intuitive interface so users can quickly understand travel risks.

Challenges Faced

Ensuring fast asynchronous data fetching without blocking the UI.
Structuring Gemini prompts for short, concise, and meaningful alerts.
Integrating translation and TTS while keeping everything responsive.
Designing a clean, interactive UI that handles dynamic content like autocomplete and live alerts.