holding

What the company appears to be:

The project "holding" is described as a personal immigration strategist for O-1A/EB-1A visa applicants. It allows users to input structured facts about their case and evaluates those against USCIS's own rules, providing verdicts that link directly to official government sources.

What changed:

The author states they built this tool in response to the lack of clarity in how USCIS judges cases, particularly for high-skilled immigrants. The project is presented as an independent decision layer that avoids giving fake confidence scores and instead offers verifiable rulings based on official documents.

Single most important open question:

Is there any evidence of user engagement or adoption beyond the author’s own development efforts? The description does not indicate whether real users have tested or engaged with the product, nor if it has been used in actual immigration cases.

Note

This analysis is based solely on the self-reported and unverified project description provided by the caller. No external data, traction, revenue, or customer information is available.

The description states:

• "holding is a personal immigration strategist for O-1A and EB-1A."
• Users enter structured facts about their case.
• It evaluates those facts against USCIS's own rules.
• Each verdict links directly to the official government rule cited, down to paragraph and hash of source text.
• It refuses to generate a confidence score or fake number.
• It is bilingual, Chinese-first for its first market.
• It runs locally using transformers.js with no API key.

Inference: The tool appears to be a structured legal evaluation engine that uses AI to analyze immigration criteria and present findings tied to official USCIS documents. However, the actual functionality and interface are not described beyond this high-level structure.

The description states:

• The product aims to be an "independent decision layer for high-skilled immigration."
• It positions itself as a neutral reference that avoids giving fake confidence scores.
• It contrasts with lawyers, agencies, and chatbots by refusing to dilute judgments into made-up numbers.
• It is not intended to replace lawyers but to prepare applicants for legal representation.

Claim: The author claims the tool is built on structural neutrality and honesty — avoiding scoring systems that could mislead users.

Inference: This positioning reflects a strategic differentiation from competitors who offer confidence scores, which may be seen as a key selling point in a market where trust is critical.

The description states:

• The tool targets O-1A and EB-1A visa applicants.
• It is bilingual, Chinese-first for its first market.
• It is designed to help people understand how USCIS judges cases without relying on lawyers or chatbots.

Claim: The target customer is high-skilled immigrants applying for O-1A/EB-1A visas.

Inference: The ICP likely includes individuals who are self-representing, particularly those from Chinese-speaking communities, due to the bilingual focus and emphasis on clarity over marketing fluff.

The description states:

• "It is not the lawyer."
• "It prepares you for the lawyer."
• "It never sells you a petition."
• "It refuses to fake a number."

Claim: The business model does not involve selling petitions or legal services directly.

Inference: There is no evidence of pricing, monetization, or revenue streams beyond the author’s own development effort. The tool is described as a preparation tool for lawyers rather than a service provider.

The description states:

• Built with Codex running on GPT-5.6.
• Uses BM25, dense, structure-only, and combined retrieval methods.
• Retrieves from an immutable 5,701-decision official corpus.
• Runs locally using transformers.js with no API key.
• Citations are verified via hash, document checksum, PDF page, and exact English span.
• Claims that fail verification are suppressed rather than replaced by summaries.
• A permutation test proves outcomes are not part of ranking.
• The system is shipped as static fixtures for local execution.

Claim: The tool uses deterministic retrieval and citation validation to ensure accuracy.

Inference: Technical rigor is emphasized, especially around data integrity and source verification. However, no evidence of scalability, performance metrics, or production deployment is provided.

The description states:

• It was submitted to the OpenAI 2026 hackathon.
• The author built it alone (team size: 1).
• No mention of users, customers, or usage data beyond personal development.

Claim: There is no evidence of traction, adoption, or user feedback.

Inference: The project appears to be in early-stage development with no external validation or real-world application yet.

The description states:

• Competitors include lawyers, agencies, and chatbots like ChatGPT.
• These tools often provide fake confidence scores.
• The author's tool refuses to do so, positioning itself as more honest.

Claim: The competitive landscape includes traditional legal service providers and AI assistants that offer unverifiable predictions.

Inference: The tool differentiates itself through honesty and verifiability, but there is no evidence of market analysis or competitor pricing or reach.

• No user engagement or adoption — the project is described only as a personal development effort.
• Unverifiable claims about accuracy or impact — while technical details are given, there's no independent validation of results.
• Legal risk: The tool may tread into unauthorized practice of law territory, especially if it gives legal advice or interpretations.
• Limited scalability — the system runs locally and uses static fixtures, suggesting limited production readiness.
• No revenue model — no indication of how this would be monetized or whether it has any commercial viability.

Inference: The project lacks commercial traction, user feedback, or clear path to monetization. Its legal standing is unclear, and its technical implementation may not scale beyond prototype status.

1. Has the tool been tested with real users? What were their reactions?
2. How does it handle edge cases or ambiguous legal standards?
3. Is there any risk of being considered an unauthorized practice of law?
4. Are there plans to expand beyond O-1A/EB-1A visas?
5. What is the plan for validation by legal professionals?
6. How is the corpus of USCIS decisions maintained and updated?
7. Can the tool be used in other jurisdictions or immigration types?