Coral Collective AI

Coral Collective AI is a mobile application concept that enables divers to bulk upload reef photos, which are analyzed by an AI and machine learning pipeline to identify coral species and assess reef health. The processed data feeds a community platform that connects citizen scientists to professional reef researchers and conservation organizations.

The product was conceived, defined, designed, and demoed in three days at Purple Maiʻa Foundation's Purple Prize hackathon. It placed first. This case study is about the product decisions that won. Specifically: the architectural decision to treat community engagement not as a feature layer but as the product's core structure.

Coral reef monitoring in Hawaiʻi and across the Pacific is labor-intensive and resource-constrained. Professional reef surveys are expensive, infrequent, and cover a fraction of the reef systems that need monitoring. Meanwhile, hundreds of recreational divers are in the water every week with cameras, capturing data that never gets systematically analyzed.

The gap isn't data. It's infrastructure. There's no pipeline that connects what recreational divers see to what reef scientists need, and no platform that gives divers a reason to contribute their footage beyond posting it to social media. That's the product problem Coral Collective AI was built to solve.

In a 3-day sprint with a cross-functional team, the product leadership role isn't a title. It belongs to whoever holds the threads together and makes the calls that keep the build coherent. That was this role.

• Defined the product vision and the core architectural argument: community engagement as infrastructure, not marketing
• Owned the product roadmap for the sprint: determining what gets built in three days and what gets scoped to a future phase
• Served as the cross-functional connector, keeping developers, designers, and the pitch narrative in alignment under time pressure
• Made the call on Salesforce integration: meeting conservation organizations where their data infrastructure already lives rather than asking them to adopt a new system
• Defined how AI pipeline outputs are presented to non-expert users, solving the core UX problem: making machine-generated reef health data legible to a recreational diver
• Defined the circular design architecture: structuring the product to share data resources with adjacent environmental monitoring projects

The decision that won the Purple Prize wasn't technical. It was architectural, and it started with a product question, not a technical one:

• Community engagement as product architecture, not a feature. The AI species identification pipeline is only as good as the data that feeds it. Building the community interface as a core structural layer, not an add-on, meant the product had a self-sustaining data model built in from the start
• Salesforce integration from day one. Conservation organizations already run on Salesforce. Asking them to adopt a new data system would have killed adoption before launch. Meeting them where their infrastructure lives was the right scope call
• Circular design architecture. The product was structured to share reef monitoring resources with adjacent environmental projects, making Coral Collective AI a node in a larger conservation data network rather than an isolated application
• AI output legibility as a first-class product requirement. The species identification data is scientifically meaningful but not intuitively readable. Designing the presentation layer for a recreational diver, not a marine biologist, was a deliberate product decision, made before a single screen was designed

Coral Collective AI placed first at Purple Maiʻa Foundation's Purple Prize 2024, a competitive 3-day hackathon. The winning argument was architectural: treating community engagement as a structural product requirement rather than a growth strategy.

The product concept establishes a replicable model for citizen science platforms that need to bridge recreational participation with professional research pipelines. The architecture decisions are directly applicable to any conservation tech product that depends on community-contributed data.