Deterministic extraction, not AI guesswork

Our Mission

Making important documents understandable, so you know exactly what you're agreeing to before you sign.

We believe that navigating important life decisions shouldn't require a law degree. Whether you're reviewing a contract, signing a lease, starting a new job, or checking an insurance policy, you deserve clarity about what you're agreeing to.

Our Company

DocAssessment is built and operated by Macpherson Publishing, based in Ontario, Canada.

What We Do

DocAssessment analyzes legal documents and shows what they say. We extract key information and flag potential risks, with the wording each flag came from where it can be quoted.

• Document Analysis: find dates, parties, and money amounts in your document
• Risk Detection: flag clauses that may deserve a closer look, such as automatic renewal, non-compete terms, or arbitration
• Questions Worth Asking: a short list of questions to put to the other party before you sign

How It Works

DocAssessment reads your document with pattern matching and fixed rules. No language model is involved.

• Deterministic Extraction: pattern matching, natural language processing, and reference libraries extract facts from your document, and a fixed scoring formula builds the risk score, adding points for a clause only when its own heading names a risk category. No language model reads your document
• Transparency: where a finding is tied to specific wording, it cites the exact source text from your document

Because no language model writes any part of the report, it cannot contain an AI hallucination. Pattern matching can still misread a passage, so check each flag against the wording it points to.

Our Methodology

Most AI document tools send your raw document directly to a language model and ask it to summarize or interpret. This produces confident-sounding output that may contain invented facts, a problem known as hallucination.

DocAssessment takes the opposite approach: no language model reads your document. Here is the exact sequence:

1. Text extraction from your document (PDF parsing, OCR for images, Word document parsing)
2. Keyword-based document type detection (no AI)
3. Pattern-based clause and section segmentation (no AI)
4. Fact extraction using regex and NLP libraries: dates, parties, amounts, deadlines (no AI)
5. Heuristic risk scoring based on extracted clause tokens (no AI)
6. Report building: each finding is shown with the passage it came from where that passage can be quoted whole

The result: analysis you can verify and, where a flag is tied to specific wording, a risk flag you can trace to that text.

Privacy First

We take your privacy seriously:

• Two-Tier Retention: source documents auto-deleted at 24 hours; structured analysis kept for 30 days, then fully removed
• No AI Training: we never train AI models on your uploaded documents
• Anonymous Usage: you can analyze documents without creating an account
• Encryption at Rest: stored documents are encrypted for security

Trust & Transparency

DocAssessment is an information tool, not legal advice.

DocAssessment provides factual analysis to help you understand your documents. It does not interpret the law, advise on legal strategy, or tell you what to do. For important legal decisions, consult a qualified attorney.

Our goal is to give you the clarity you need to have more productive conversations with legal professionals, not to replace them.

Read our full disclaimer