852Planner

Home page
Question 1 page
Question 2 page
Question 3 page
Question 4 page
Question 5 page
Question 6 page
Question 7 page
Trip Overview page
Loading page
Itinerary page 1
Itinerary page 2
Expense Breakdown page
Sign Up page
My Dashboard page

Inspiration

Our team consists of two people: a Hong Kong student, Ethan, and a Korean student, Hyun, both studying at HKUST. When Hyun first came to Hong Kong to travel for a few days with his family, he only saw the typical tourist spots. But after living in Hong Kong, Ethan showed him around to reveal the real local attractions, and Hyun discovered epic new places for hiking, island hopping, and much more that he had missed on his previous trip. We believe there's so much more to Hong Kong than the beautiful city views, sea, and shopping. Hong Kong has amazing hiking trails, nature, and beaches that haven't been fully discovered by visitors yet. We wanted to provide that local and varied experience, using AI to add a smoother analytical approach for better planning.

What it does

After a short survey on your basic trip information and preferences, we create a detailed plan for your travels in Hong Kong. We display the daily weather forecast and suggest ongoing events or festivals during your stay. With visual maps and routes, plus short descriptions of each attraction or restaurant, you can already imagine what your trip will be like—and it's made just for you. We also provide budget estimates for each destination and a total expense breakdown so you can plan ahead. You can press the navigate button to jump to Google Maps with the destination path selected automatically. You can save it and share it with your travel mates. This all takes less than five minutes. Compared to searching on Google, blogs, or even ChatGPT, it's a much faster and more efficient way to plan your trip in Hong Kong. We believe our product will help tourists get more out of their Hong Kong travels and, for Hong Kong itself, showcase more of its unique, lesser-known attractions to stimulate growth in the industry.

Business Value & Use case

Our product doesn't generate revenue on its own, but that doesn't mean we don't fit the market well. In the Hong Kong Tourism Board's 2025-26 work plan, they emphasize implementing the Development Blueprint for Hong Kong’s Tourism Industry 2.0, focusing on the concept of “tourism is everywhere” to encourage repeat visits, launching smart tourism tools to enhance visitor experiences and enable the travel trade to develop new products, and bundling mega events with always-on experiences for thematic promotions. They also promote green tourism through the year-round “Hong Kong Great Outdoors” platform, highlighting hiking trails, islands, beaches, and nature reserves to showcase the city's diverse natural attractions. We create business value by helping tourists have a richer, more local experience during their travels, stimulating Hong Kong's tourism industry overall. To expand our service and reach real users, we need support, which ties into our future development plans to collaborate with the Hong Kong Tourism Board.

How we built it

We used Amazon Q Developer to develop the whole UI interface, fix bugs, and implement APIs for our application. We leveraged it by creating fast and developing a MVP in just a few days. It helped a lot in fixing minor UI bugs and formatting fetched API data for our app. We extensively used Amazon Q Developer—so much so that 90% of our code was generated by it.

Challenges we ran into

We wanted our 852Planner to stand out from other AI planners already in the market. We aimed to generate more local and specific data just for Hong Kong. Our first idea was to train a RAG with data we collected about attractions and restaurants in Hong Kong. We failed this approach because the data wasn't tagged and refined well. This was a tough experience, as we spent a long time collecting data from the government's database and attempting other sources like TripAdvisor, Foodpanda, and more (which we couldn't do due to their website policies). The final approach we landed on was integrating Gemini with calls to the Google Places API to search and collect images and data for your tailored itinerary. Thankfully, this worked with valid and trustworthy data from Google Places.

Accomplishments that we're proud of

We really focused on polishing the UI and UX so it's fully usable, even though it's just a demo. We believe we fully accomplished that goal by beta-testing with 10+ people and iterating based on their valuable feedback. If you check our commits, you'll see we kept adding slight UI adjustments and new features from the feedback we got. We're proud that we have a final product that's truly polished. And one thing we really think was a good decision is creating the MVP (Minimum Viable Product) super fast. After we had the idea, we built the MVP in just a few days. From there, we tested with people and iterated so many times we can't even count.

What we learned

The most important thing we learned was to create an MVP fast, keep beta-testing with actual people, and keep iterating. Our first MVP was just a few survey questions with a text-generated itinerary listing only destinations, which captured the core idea of an AI-powered itinerary generator. From there, we iterated a ton to get to the polished product we have now. This approach worked incredibly well. Another thing is that AI really codes well, and we should leverage it as much as possible.

What's next for 852Planner

We currently have a fully working AI-powered itinerary generator. We'd like to improve the AI solution with a better and more precise approach to gather local restaurants and attractions that aren't on Google Maps. Our original approach integrates Google Maps with Gemini to search for good attractions and restaurants, but we believe we can improve it much further. We'd like to keep enhancing the user experience with necessary features that travelers might need. For example, using the MTR and bus real-time API, we can provide arrival times for travelers in our web app. We want to create blog posts about unique local places that tourists must visit. This can help make unknown spots more public and stimulate tourism in Hong Kong. We'd also like to implement features to avoid crowded attractions using the government API on crowd data. Our final goal is to propose this app to the Hong Kong Tourism Board for collaboration on collecting better data to enhance user experience. And finally, publish it as an app on the Google Play Store and Apple App Store.