PackWise

Inspiration

PackWise was inspired by a problem almost everyone faces, but nobody really solves well: packing is stressful, inefficient, and surprisingly complicated. Whether you are flying for vacation, going on a group trip, road-tripping, or moving, people are forced to juggle airline rules, weather, activities, luggage sizes, documents, and limited space, usually with nothing more than screenshots, notes apps, and guesswork.

We realized this was not just a checklist problem. It was a reasoning problem. People do not only need reminders; they need help deciding what to bring, what not to bring, what will fit, what should go where, and what someone else in the group can bring instead. That pushed us to build PackWise as an AI-powered packing strategist rather than just another travel planner.

What it does

PackWise is an AI travel and moving copilot that helps users pack smarter, faster, and with less stress.

Users can create trips, enter details like destination, duration, luggage size, and travel type, and upload photos of the items they already own. PackWise then uses AI to identify those items, estimate what they are, and reason about what is missing, what is unnecessary, and how everything should be packed. PackWise works as a phone app and website.

It helps users:

detect missing essentials,
understand destination-specific needs,
organize items between backpack, carry-on, checked bag, or moving vehicle,
optimize space,
decide whether to wear, pack, or buy items at the destination,
manage passport and travel document reminders,
coordinate group packing so people do not duplicate shared items,
and even account for practical constraints like how much a traveler can comfortably carry.

In short, PackWise turns chaotic packing into a personalized, optimized strategy.

How we built it

We built PackWise as a cross-platform app in Flutter so we could support both mobile and web from a single codebase.

For authentication and persistence, we integrated Firebase Auth and Firestore. We created a custom auth flow with Google sign-in and guest mode, then used Provider-based state management through a centralized AppState to track users, trips, lists, and UI state.

On top of that, we built an AI workflow powered by Google Gemini. Users can take photos of their items using the camera or image picker, and Gemini Vision converts those images into structured item data. We then run multiple specialized Gemini prompts on top of that shared item context to handle things like:

missing-item detection,
bag placement optimization,
destination-aware packing suggestions,
pack-vs-buy reasoning,
and space optimization.

For moving and road trip flows, we also let users enter vehicle dimensions and use those values to estimate available volume. That way, our AI is not only reasoning about what to bring, but also about whether it can realistically fit into the user’s suitcase, car, van, or truck.

Challenges we ran into

One of our biggest challenges was balancing intelligence with speed. Because our app runs several AI analyses rather than a single prompt, we had to consider latency and user experience carefully. We solved this by parallelizing model calls where possible and surfacing clear loading states so the app still felt responsive.

Another challenge was getting reliable, structured outputs from the model. Since our UI depends on structured item data, we spent a lot of time refining prompts, schemas, and parsing logic so Gemini would consistently return usable results.

Finally, managing the scope was a challenge. While we had many ideas for features, we prioritized building a solid vertical slice during the hackathon instead of attempting to implement everything at once.

Accomplishments that we're proud of

We are proud that PackWise feels like a real product, not just a model demo.

We built a working experience where users can:

sign in,
create and manage trips,
upload photos of items they already own,
get AI-powered packing analysis,
and receive actionable recommendations that actually make travel and moving easier.

We are especially proud of how PackWise combines multiple layers of reasoning in one product: image understanding, destination awareness, airline and luggage constraints, group coordination, and space optimization.

We are also proud that the app goes beyond a generic packing list. It acts more like a logistics copilot that helps users make better decisions, reduce waste, and lower stress before a trip even begins.

What we learned

We learned that the most valuable AI products are not the ones that simply generate text; they are the ones that fit naturally into a real workflow.

We also learned that orchestration is just as important as the model itself. Turning an image into useful structured data, then using that data across several layers of reasoning, was one of the most important technical insights from this project.

On the product side, we learned how important resilience and UX are. Handling slow authentication, slow database loads, and multi-step AI reasoning gracefully was just as important as building the intelligence itself.

Most of all, we learned that solving a real, relatable problem with thoughtful execution is much more powerful than building something flashy but shallow.

What's next for PackWise

Next, we want to make PackWise even more personalized and visually intelligent.

We want to add:

more accurate size and weight estimation from photos,
richer airline-specific restriction checking,
visual suitcase and vehicle layout maps,
stronger group-trip coordination,
smarter outfit planning,
and real-time travel prep timelines tied to departure dates.

We also want to improve the moving flow so PackWise can become not just a travel packing assistant, but a full space-aware logistics platform for both travel and relocation.

Our long-term vision is for PackWise to become the system people trust whenever they need to figure out what to bring, how to fit it, and how to travel smarter with less stress.