ResolveAI

What the company appears to be: ResolveAI is a self-reported AI-powered tool that processes user-uploaded receipts, screenshots, photos, and PDFs into structured resolution workflows for common disputes (e.g., refunds, subscription issues, customer complaints). It claims to generate evidence maps, next actions, editable messages, and progress tracking without automatically sending messages or contacting third parties.

What changed: The project was built as part of an OpenAI 2026 hackathon submission. It is described as a complete, responsive product with three end-to-end workflows, multimodal support, defensive validation, and deterministic test coverage. No prior version or evolution is mentioned.

Single most important open question: Is there any evidence that users actually engage with the tool beyond its initial demo state? The description lacks any indication of real-world usage, adoption, or traction.

The description states that ResolveAI is a Next.js 15 and React 19 application written in strict TypeScript. It supports three specific dispute resolution scenarios:

• Damaged products and refund or replacement requests;
• Unexpected charges and subscription disputes;
• Customer complaints that small businesses need to resolve.

It allows users to describe what happened, optionally attach files (receipts, screenshots, photos, PDFs), and returns:

• An evidence map separating supported facts from gaps;
• Important dates, uncertainties, and an evidence-readiness score;
• Prioritized next actions with progress tracking;
• A recipient-email result that never invents an address;
• One editable message the user reviews before taking action;
• Controlled message refinements and up to five recent browser-saved cases.

The system does not contact anyone, change accounts, or send messages automatically. Raw uploaded files are not persisted by the application.

Inference: The product is a frontend-heavy, AI-assisted workflow tool designed for small-scale dispute resolution, built using modern web stack with multimodal inference and structured outputs.

The author states that ResolveAI was inspired by everyday disputes that are too small to justify professional help but confusing enough to delay action. It aims to turn scattered evidence into a clear, user-controlled resolution workflow.

Claim: The tool turns unstructured data (receipts, screenshots, etc.) into actionable steps and editable messages.

Inference: This is positioned as a consumer-facing assistant for small disputes, not a B2B platform or enterprise solution. It emphasizes control and non-invasiveness over automation.

The description states that ResolveAI supports three focused scenarios involving:

• Damaged products and refund or replacement requests;
• Unexpected charges and subscription disputes;
• Customer complaints that small businesses need to resolve.

It is implied that the primary users are individuals who encounter such issues, possibly with some involvement from small business owners or customer service teams.

Inference: The target user appears to be individuals dealing with common consumer disputes, potentially including small business stakeholders. No explicit segmentation beyond these use cases is provided.

There is no mention of pricing, monetization strategy, or business model in the description.

Not evidenced.

The application is built using:

• Next.js 15, React 19
• TypeScript, Zod schemas, Tailwind CSS
• Integrates with:
  • NVIDIA NIM (nemotron-nano-12b-vl)
  • Google Gemini
  • OpenAI API
• Uses a two-pass vision flow for image processing to ensure grounded structured output
• Implements server-side validation, file signature checks, and request limits
• Includes 103 offline tests, strict TypeScript validation, zero-warning linting, and production build

The system uses deterministic mock mode for demos without requiring API keys.

Inference: The tool is technically robust, with a focus on security, reliability, and structured output. It avoids open-ended AI chat in favor of controlled workflows.

There is no evidence of traction, revenue, customers, or usage beyond the initial hackathon submission.

The description states that it's a complete responsive product and includes:

• 103 passing tests
• Strict TypeScript validation
• Zero-warning linting
• Production build
• Public deployment

However, these are development maturity indicators, not signs of user adoption or market traction.

Not evidenced.

No mention of competitors or competitive landscape is provided in the description.

Not evidenced.

• No evidence of real-world usage: The tool is described as a hackathon submission with no indication of actual users.
• Self-reported maturity: All claims about functionality, testing, and delivery are from the author’s own account.
• Limited scope: Only three scenarios are supported; no expansion plans beyond that are detailed.
• No monetization strategy: No evidence of how this would be commercialized or scaled.
• High technical dependency on AI providers: Reliance on NVIDIA NIM, Google Gemini, and OpenAI APIs may pose risks if those services change or become unavailable.

1. Has the tool been tested with real users beyond the demo environment?
2. What is the intended monetization model for this product?
3. How does the tool plan to scale beyond the current three use cases?
4. Are there any plans to integrate with existing dispute resolution platforms or services?
5. What are the long-term implications of relying on external AI providers for core functionality?