The Historian

Experience spatial-temporal evolution of iconic global landmarks using multimodal and data augmentation, providing a learning experience of historical reconstructions and AI-narrated histories.

Comment

Inspiration

The app was inspired by my travel journeys. I wanted to build an app that would allow people to immerse into the history of places of interests, making the learning experience simple, interesting and fun for everyone. The goal was to experiment with Gemini 3 API to build and enhance the user experience of learning history.

What it does

The Historian is a learning app whereby users can upload photos taken of landmarks or places during their travel or use the Scholar-History Agent to record an audio prompt and ask the app to do research of the landmark and produce interesting facts using pictures and audio narration, allowing users to immerse and experience visual 360 panorama of the landmark. Users can step back in time using a timeline slider and see how the landmark looked in the past and learn about it by listening, rather than just reading. All history research discoveries are saved in the client cache and can be backed up and restored on another client browser. This app is usable by everyone including school children.

How we built it

The app was built initially using Google AI Studio and vibe-coding. The app was created using TypeScript, React framework and Gemini 3 API multimodal capabilities to generate images, search grounding and Gemini 2.5 Flash TTS to convert text to audio. However, I later switched from Google AI Studio to Google Antigravity.

Challenges we ran into

Google AI Studio was working for me up to a certain point when developing the app. Sometimes AI Studio would apply a fix and also change the entire look and feel of the app. AI Studio would overwrite fixes when it fixed another issue. I found it difficult to keep track of the changes made in AI Studio. The deployment process to Google Cloud was also not transparent and I could not see what or how the code was deployed. When I switched to Google Antigravity I felt I had more control and understanding of the code change and the deployment process made by the agent by referring to the walkthrough.md and tasks.md

Accomplishments that we're proud of

To be able to prototype, develop and deploy the app in days rather than weeks.

What we learned

I have a better understanding of Gemini 3 multimodal capabilities. I learned how to do deployment to Google Cloud. It is possible to build an app in a team of one using vibe coding and Google AI Studio and Google Antigravity. In the past, you need someone with UI/UX skill to do the front-end development and someone else with DevOps to do the deployment. Now, the roles can be combined with the use of code assistance.

What's next for The Historian

The app can be generalized as a template for learning any subject. The app can be enhanced to include guardrails to disallow sensitive information and inappropriate prompts. Adding a login page for user registration and to keep a user database.

Built With

Share this project:

Updates