ScamShield Pro

Note: All content is based on a self-reported project description. No external verification, revenue data, customer names, or independent sources are available.

The description states that ScamShield Pro is a local-first Chrome extension that performs multiple layers of scam detection:

• Text-based detection: Uses a fine-tuned DistilBERT model to classify text as Safe, Risky, or Scam.
• Image OCR support: Allows users to drag-select text from images on webpages and analyze it locally using Tesseract.js.
• Link inspection: Detects lookalike domains, URL shorteners, punycode, hidden redirects, and visible-link mismatches.
• Anti-phishing scan: Checks for fake login pages by inspecting claimed brand, current domain, password fields, and form destinations.
• Company research feature: Enables manual or selected company name input to return public evidence with source links (via Tavily API and Node.js backend).

It uses:

• Chrome Manifest V3
• ONNX Runtime Web with WASM
• Transformers.js
• Tesseract.js
• A Node.js backend hosted on Render for optional company research

Inference: The product is described as a browser extension, not a SaaS platform or API service. It is built to run locally and does not appear to be a commercial product yet.

The author claims ScamShield Pro addresses the growing sophistication of scams, including:

• Fake job offers
• Phishing emails
• Fake login pages
• Shortened links
• Screenshot-based recruitment scams

They state that people should not need to copy content into a chatbot to assess risk — instead, they want a fast safety layer directly inside the browser.

The positioning is local-first security, emphasizing:

• Privacy (no data leaves the browser)
• Speed (real-time detection)
• Explainability (shows why something was flagged)

Claim vs Fact: The author positions this as a solution to modern scam trends, but no evidence of actual user feedback or market validation is provided.

The description states that ScamShield Pro targets:

• Students
• Job seekers
• Everyday internet users

These are described as individuals who are increasingly fooled by professional-looking scams.

Inference: The target audience appears to be general consumers rather than enterprise clients. No segmentation beyond “internet users” is evident.

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

Not evidenced

Key technical elements:

• Uses local inference via ONNX Runtime Web and Transformers.js
• Implements Tesseract OCR for image text extraction
• Runs on Chrome Manifest V3
• Includes an offscreen document for model execution
• Has a Node.js backend hosted on Render for optional company research
• Integrates with APIs like Gemini, Tavily Search, and Codex/GPT-5.6

Inference: The architecture shows a focus on privacy, performance, and integration complexity — but no evidence of deployment scale or production readiness.

The project is described as a hackathon submission (OpenAI 2026) and was built by one person (Rehan Raza). There is no evidence of:

• Users
• Revenue
• Customers
• Adoption metrics
• Product-market fit
• Commercial traction

Not evidenced

No competitive analysis or mention of existing tools is included in the description.

Not evidenced

• Single-person development: The entire project was built by one individual, raising questions about scalability and long-term maintenance.
• No commercial traction: No evidence of users, revenue, or market validation.
• Local-first design may limit impact: While privacy-focused, local execution could reduce effectiveness if not widely adopted.
• Dependency on AI accuracy: The system relies heavily on ML models that may miss nuanced scams or generate false positives.
• Limited scope: The tool is focused on browser-based detection and lacks broader ecosystem integration.

Inference: The lack of any commercial or user-facing data makes it difficult to assess viability as a product or business.

1. What is the current usage rate or adoption level among users?
2. How do you plan to monetize this extension beyond its current open-source/hackathon state?
3. Have you tested the model’s accuracy in real-world scenarios with actual scam content?
4. Are there any plans for expanding beyond Chrome or adding mobile support?
5. What are your long-term goals for product development and market positioning?