ReelTank

1. Inspiration

ReelTank began with a simple desire to preserve everyday memories.

For many anglers, the story of a catch ends with a photo posted on social media. The picture may survive, but the more meaningful details often disappear: What species was it? Where was it found? Was it rare? Could it legally be kept?

I wanted to turn that brief moment into something more memorable. Instead of treating a catch as just another photo, ReelTank transforms it into a discovery—part field guide, part treasure hunt, and part personal aquarium.

2. What it does

ReelTank turns fishing into an interactive collecting experience.

In the intended experience, an angler uploads a photo of a catch. The app combines visual characteristics—such as body shape, markings, mouth shape, and fin placement—with location data to help identify the species.

It then presents useful information about the fish, including:

• Common and scientific names
• Identifying characteristics
• Habitat, diet, and behavior
• Handling and safety notes
• Location-aware fishing regulation guidance
• A link to official regulations before the fish is kept

Users can give each fish a name and add it to a personal digital aquarium. As their collection grows, they unlock badges for discovering new species, visiting different fishing locations, and reaching collection milestones.

The three-day prototype uses curated Florida freshwater sample catches instead of live photo recognition. Each sample includes its own location metadata, allowing the demo to communicate the complete product experience without pretending that a production AI system is already running.

3. How we built it

I first mapped the core journey in Figma:

Select a catch → identify possible species → review fish information → name the fish → add it to the aquarium → unlock collection progress

I then built ReelTank as a native iOS prototype using:

• SwiftUI for the interface, navigation, interactive states, and aquarium animation
• SwiftData for storing named fish, species, locations, dates, and catch records locally
• Figma for interaction design, visual direction, and the American retro fishing identity
• A protocol-based recognition layer that keeps the identification experience separate from the interface
• A deterministic local demo service that returns reliable candidate species for bundled sample catches
• Curated Florida freshwater data for six representative species, including habitat, behavior, identification features, handling notes, and fishing guidance
• Bundled sample metadata for Lake Tohopekaliga, Lake Okeechobee, and Lake Kissimmee
• Local achievement logic for collection milestones, badge progress, and reward states

The aquarium uses lightweight SwiftUI animation to move saved fish through the scene. Catch records remain on the device, so the prototype requires no login, backend, API key, or network connection.

AI-assisted coding tools helped me translate the product design into a working application, diagnose build issues, and iterate quickly. However, the product decisions, interaction structure, visual system, content hierarchy, and final scope were guided by my design judgment.

4. Challenges we ran into

The biggest challenge was scope.

I am a designer, this was my first hackathon, and it was also my first time seriously experimenting with vibe coding. I had to learn how to translate design intent into application logic while making decisions quickly enough to finish a working prototype.

The original concept depended on users uploading their own photographs and using AI to identify arbitrary fish species. Building and validating that experience would require a reliable vision service, permission handling, backend security, privacy decisions, error states, and extensive testing across many species and photo conditions.

With only three days and limited resources, implementing that responsibly was not realistic. Instead of presenting an unreliable feature as complete, I created a guided demo using bundled Florida freshwater catches and deterministic local results.

Another challenge was presenting fishing regulations responsibly. Rules can change by season and waterbody, so the prototype clearly labels its content as guidance and directs users to official Florida FWC regulations before making a decision.

5. Accomplishments that we're proud of

I am proud that ReelTank became a functional native iOS prototype rather than remaining only a design concept.

Within the hackathon timeframe, I completed an end-to-end experience that allows users to:

• Select a sample catch with location metadata
• Review possible fish species
• Explore detailed species information
• Name a fish
• Save it locally
• Watch it appear in a personal aquarium
• Reopen its collection record
• Make progress toward achievement badges

I am also proud of the product’s emotional identity. The vintage American fishing-club style, illustrated field-guide language, tactile badges, and personal aquarium make the experience feel like an outdoor treasure hunt rather than a clinical identification utility.

Most importantly, I found a way to communicate the complete vision honestly without pretending that unfinished AI functionality was production-ready.

6. What we learned

This project taught me that successful hackathon work is not about building every possible feature. It is about identifying the smallest experience that still communicates the heart of the product.

I learned how to:

• Convert a user journey into SwiftUI screens, navigation, state, and persistent data
• Design a technical structure that can later replace demo logic with a real service
• Use AI-assisted development as a collaborator while continuing to make independent product decisions
• Test and debug a native iOS application
• Scope ambitious features around time, reliability, privacy, and user trust
• Treat uncertainty and legal information as part of the user experience rather than hiding them

Most of all, I learned that design and code are not separate stages. Working directly with both allowed me to revise the product much faster and make every interaction more intentional.

7. What's next for ReelTank

The next step is to turn the guided prototype into a real-world discovery tool.

Planned improvements include:

• Camera and photo-library uploads
• Production-quality visual species recognition
• Location-assisted identification
• Secure server-side AI processing
• Confidence scores and alternative species suggestions
• More accurate, waterbody-specific regulation guidance
• GPS-based fishing spot check-ins and a personal discovery map
• More Florida freshwater species
• Expansion into additional aquatic and outdoor species
• More collection milestones, badge families, and seasonal challenges
• Improved aquarium environments and fish-specific movement
• Optional accounts, cloud backup, and collection syncing

The long-term vision is for ReelTank to become a personal field journal for outdoor discoveries—one that helps people understand what they find, remember where they found it, and turn every trip into a new chapter of an ongoing treasure hunt.

Built With

• codex
• gpt5.6
• ios
• swiftdata
• swiftui