AI Next-Step

• 

  AI Next-Step: The smart suggestion screen in action.

Inspiration

When I started I went to Google AI Studio to know about the hackathon. From there I got to know about tracks and the three services Bolt provide. And from the tracks I felt like I was most comfortable in user experience. Then I started with what I want to build and AI gave me ideas. One of the ideas stuck with me but it was not to my satisfaction so, I made some changes in the idea and the project became THE PROACTIVE TRAVEL ASSISTANT. But it was too complex for a beginner and time-taking, and I didn't have much time. It was a good idea, means I really liked it and especially for an oblivious person like me, that assistant would have been a blessing. So we made some changes here and there and I finally landed on AI Next-Step. It is also a great idea that connects the whole ecosystem of Bolt so we went ahead with this.

What it does

AI Next-Step is an intelligent feature for the Bolt app. After your ride ends, it analyzes where you were dropped off and the time of day. Based on this, it gives you a smart, contextual suggestion for your next move. If you're dropped off at a park, it suggests finding a nearby Bolt Scooter to explore. If you arrive downtown at noon, it suggests ordering lunch from a nearby restaurant on Bolt Food . It connects the end of a ride to Bolt's other services.

How we built it

This project was built from scratch in four days, focusing on a working prototype. Day 1: The Screen: We built the user interface with HTML and CSS to look like the Bolt app's end-of-ride screen. Day 2: The Brain: We wrote the JavaScript to call the Google Gemini AI using an API key. We engineered a specific prompt to get the smart suggestions back. Day 3: The Connection: We connected the "Brain" to the "Screen." We made it so that clicking buttons on the page would call the AI and then display the suggestion on the screen. Day 4: The Demo: We recorded a video of the working prototype and prepared the project for submission.

Challenges we ran into

Well, writing the codes was not the hard part since AI was doing it for me, and also I know HTML and CSS so it was no problem. But afterwork things like making links, GitHub Pages, and especially the video was rather time consuming and new to me.

Accomplishments that we're proud of

I am proud that I completed this project.

What we learned

Well, I got to know video editing on Canva and how I can make links of my video through YouTube without making it public. I created my GitHub account, learned how to create a repository on that, and make a link of the site. I also used the API key which I created long ago but never used it.

What's next for AI Next-Step

I feel like if Bolt likes this idea, they can really build the AI and connect all three of their services

Built With

• canva
• css
• geminiapi
• github
• googleaistudio
• html5
• javascript