TapCheck

The description states that TapCheck is a web application that:

• Takes a ZIP code as input
• Generates a plain-language water quality report in ~10 seconds
• Uses live EPA SDWIS data
• Provides letter grades and summaries for water systems
• Shows per-contaminant cards with health effects and legal limits
• Offers filter recommendations (activated carbon vs reverse osmosis vs "you don't need one")
• Includes dedicated tabs for fish, plants, and pets
• Has honest empty states that acknowledge when data is missing

The product is described as built end-to-end using Codex and GPT-5.6 during a Build Week hackathon.

Evidence The author's own write-up describes the functionality.

Inference The tool appears to be a data aggregation and interpretation platform, not a direct service provider or subscription business.

The description states that TapCheck was inspired by the problem of "14 pages long" EPA annual quality reports that are unreadable to average users. It positions itself as solving this by providing:

• Plain-language reporting
• Structured outputs with citations
• Health context for contaminants
• Filter recommendations
• Specialized content for non-human users (fish, plants, pets)

The claim evolution appears to be from a personal frustration with EPA reports to a tool that makes them accessible and actionable.

Evidence The author's own write-up describes the inspiration and positioning.

Inference The positioning is focused on accessibility and trustworthiness of public health data, not commercialization or monetization.

The description states that TapCheck serves:

• Residents who want to know if their tap water is safe
• Users interested in understanding what's actually in their tap water
• People concerned about contaminants like chloramine
• Pet owners and gardeners who need information about water effects on fish, plants, and animals

It also mentions "50 million Americans" served by systems with recent violations, though this is not a stated target customer but rather a context for the problem.

Evidence The author's own write-up describes the intended audience.

Inference The ICP appears to be general public users concerned about water quality, particularly those who find EPA reports inaccessible or confusing.

The description does not state any business model or pricing information. It only describes the product functionality and how it was built.

Evidence Not evidenced.

Inference Based on the description alone, there is no indication of monetization, subscriptions, or paid features.

The description states that TapCheck was built with:

• Codex (used for reverse-engineering APIs and scaffolding)
• GPT-5.6 (as interpretation layer with structured outputs)
• Next.js, React, Node.js, TypeScript
• Vercel deployment
• EPA Envirofacts API (SDWIS data)
• Aggressive caching due to federal API response times

It also mentions that the tool uses JSON schema for structured outputs and cites source records to prevent hallucination.

Evidence The author's own write-up describes the technical stack and approach.

Inference The technical approach suggests a data-driven, AI-enhanced product with strong focus on accuracy through citation.

The description states that this was a hackathon project built during Build Week. It mentions:

• Shareable report URLs
• Designed empty states
• Live deployment
• A feature nobody else has (telling you whether tap water will kill your goldfish)
• Accomplishments that were proud of: complete product, not proof of concept

However, there is no evidence of revenue, customers, or usage metrics beyond the hackathon submission.

Evidence The author's own write-up describes the project as a hackathon submission.

Inference There is no evidence of traction or maturity beyond the initial prototype.

The description does not mention any competitors or competitive landscape. It only states that "nobody else has" the feature of telling users whether their tap water will kill goldfish.

Evidence Not evidenced.

Inference The competitive context is unknown, but the unique selling point appears to be the specialized content for non-human users.

• No traction or revenue evidence: The project was submitted as a hackathon entry with no indication of adoption or monetization.
• Single founder: Only one team member listed (Dwain Prendergast).
• Unverified health claims: While the tool cites EPA records, it's unclear how it validates or verifies the health information presented.
• Dependency on public data: Reliance on EPA data that may be inconsistent or outdated.
• No business model: No indication of how the product will generate revenue or sustain itself beyond a hackathon prototype.

Evidence The author's own write-up and lack of any external validation.

Inference These are risks based on the limited evidence provided, not confirmed issues.

1. What is your plan for scaling beyond the hackathon prototype?
2. How do you intend to validate or verify the health information presented in the reports?
3. Have you considered how to monetize this product if it gains traction?
4. What are the legal implications of providing health-related information based on public data?
5. Are there any partnerships or integrations with utilities or health organizations planned?
6. How do you plan to handle updates to EPA data and ensure accuracy over time?

Evidence Not evidenced.

Inference These questions arise from the lack of evidence around traction, monetization, and risk management.