Methodology

How Revealr Analyzes Your Contract

Revealr is an AI tool, not a law firm. This page explains exactly what it does, how it works, and where its limits are — so you can use the results appropriately.

What happens when you upload a contract

You upload a contract

Drag and drop or select your file — PDF, Word (.doc/.docx), or image (JPG, PNG) up to 20 MB. Your document is encrypted in transit immediately on upload using TLS.

The AI reads the full text

Our AI model processes the complete document — every clause, every section, including footnotes and addenda. This isn't a keyword search. The model reads contractual language in context.

Risky clauses are identified and categorized

Each clause that deviates from typical standards — or that commonly causes problems for the signing party — is flagged with a severity level, a location reference (section number), a plain-English explanation, and a specific recommended action.

You get a risk score and report

A 0–100 risk score summarizes the overall finding. The free preview shows your most significant flags. The full report ($19) includes every flag, the complete action plan, a downloadable PDF, and email delivery.

What the AI is trained to detect

Revealr has been designed around the clause types that most commonly lead to disputes, financial loss, or unintended obligations for the signing party. Specific examples include:

Security deposit conditions and refundability
Early termination fees and buyout options
Automatic renewal provisions and notice windows
Landlord entry rights and inspection clauses
Non-compete scope, geography, and duration
IP assignment and moonlighting carve-outs
Termination terms: for-cause definitions and severance
Signing bonus and equity clawback provisions
Mandatory arbitration and class action waivers
NDA scope, exclusions, and duration
Indemnification scope and liability caps
Payment terms, kill fees, and revision limits

The AI also surfaces other unusual or one-sided language it encounters, even if the clause type isn't in a predefined list. The underlying model understands legal language in context, not just pattern-matching.

What the severity levels mean

Every flag is assigned one of three severity levels based on the potential impact on the signing party:

CRITICAL

Clauses that create significant financial exposure, that may violate your statutory rights, or that impose obligations substantially worse than the norm. These require attention — either negotiation before signing or legal advice.

WARNING

Clauses that are unfavorable, unusual, or one-sided in ways that may matter to you. Not necessarily dealbreakers, but worth understanding and potentially negotiating.

INFO

Clauses worth knowing about — standard provisions with non-obvious implications, terms that differ from common defaults, or conditions that may be relevant if circumstances change.

The 0–100 risk score

The risk score is a composite indicator based on the number of flags and their severity. It is designed to give you a fast orientation — not a precise legal measurement.

0–30

Low risk

Few or minor flags. Still review each one individually.

31–65

Moderate risk

Several flags worth understanding and potentially negotiating.

66–100

High risk

Multiple significant flags. Consider negotiation or professional review before signing.

A score of 0 does not mean a contract is perfect — it means the AI did not detect significant deviations from typical standards. Revealr may miss nuanced or jurisdiction-specific issues. Read the full flag list even on low-scoring documents.

What Revealr cannot do

Transparency about limitations is part of using AI responsibly. These are things Revealr genuinely cannot help with:

Provide jurisdiction-specific legal advice

Laws vary significantly by state, country, and locality. Revealr flags clauses that are commonly problematic but cannot tell you whether a specific clause is enforceable under your local law.

Evaluate the counterparty or context

Revealr reads the document. It does not know who you're dealing with, what the market standard is in your city, or whether the counterparty has a history of acting in bad faith.

Predict court outcomes

Whether an ambiguous clause would be enforced — and how — depends on facts, jurisdiction, and the specific judge or arbitrator. Revealr cannot predict how any specific dispute would be resolved.

Replace legal counsel for high-stakes decisions

For complex employment negotiations, commercial lease agreements, business acquisitions, or contracts involving significant assets, professional legal advice remains appropriate. Revealr is a first-pass, not a final opinion.

How to use your results responsibly

Step 1 — Read every flag, not just the Critical ones. Warning and Info flags often represent the difference between a negotiated outcome and a problem that surfaces 6 months into a lease or employment relationship.

Step 2 — Research flags that concern you. A CRITICAL flag is a signal to investigate further, not a guarantee the clause is illegal. Look up your state's tenant protection law, labor statutes, or contract enforcement standards for the specific clause type.

Step 3 — Use the report in your negotiation. Revealr's output gives you specific, factual starting points for negotiation — the clause number, the specific language, the issue, and a suggested alternative. This is more effective than vague concerns.

Step 4 — Consult a professional for high-stakes decisions. For agreements involving significant assets, equity, IP rights, or career-defining restrictions, use Revealr as your first pass and bring the flagged clauses to a licensed attorney in your jurisdiction. This makes professional review faster and more focused.

How your document is handled

Your document is encrypted in transit using TLS. It is stored temporarily on Vercel Blob storage only for the duration of analysis and report generation, then deleted. We do not retain copies of your documents after delivery, and we do not use your documents to train AI models.

The AI analysis is performed via Anthropic's API (Claude). Anthropic's API usage terms prohibit using API inputs for model training by default.

We recommend against uploading documents that contain information beyond what is necessary for the analysis — for example, if a contract contains trade secrets or sensitive third-party information, you may want to redact those sections before uploading.

See our full Privacy Policy for details on data handling, retention, and third-party services.

Supported file types

PDF (text-based)

Best results. Most contracts are shared as text-based PDFs.

PDF (scanned/image)

Supported via OCR. Accuracy depends on scan quality.

Word (.doc, .docx)

Fully supported. Password-protected files cannot be processed.

Images (JPG, PNG)

Supported for photographed contracts. Well-lit, legible images produce best results.

Maximum file size

20 MB per upload. Documents over 80 pages may take slightly longer.

Languages

Optimized for English-language contracts. Other languages may produce less accurate results.

See it in action

Upload any contract and get a free preview analysis in under 60 seconds.

Analyze a Contract →