Chris is an artificial intelligence system, not a person. There is no human called Chris at EquityBankLoan. The name is a label for the automated research process this site uses, so that you can tell which parts of the work were done by software.
What Chris does
Chris is EquityBankLoan’s AI research agent. Chris helps research publicly available lender websites, disclosures and product documentation; organises lender information; identifies changes and inconsistencies; and prepares research for editorial review.
In practice that means:
- Locating lender product pages, disclosures, rate pages and official FAQs
- Extracting the terms a lender publishes, with the exact wording
- Structuring those terms into a consistent record so products can be compared
- Comparing source documents against each other and against what we already hold
- Identifying which facts a lender does not publish
- Flagging inconsistencies and changes for a human to look at
- Assisting with drafts of educational material
Research produced by Chris is overseen by EquityBankLoan founder and editor Michael Yanda.
What Chris does not do
- Chris does not verify anything. Automated research prepares evidence for review; it is not itself human verification.
- Chris does not decide what gets published.
- Chris does not fill a gap with a typical figure when a lender publishes nothing. An unpublished fact is recorded as unpublished.
- Chris does not take a requirement from another publisher and attribute it to a lender who has not stated it.
- Chris does not rate, rank or recommend lenders.
- Chris does not give advice, and has no professional financial expertise, credentials or licenses.
What sources Chris uses
Primary lender sources, wherever they can be found: the lender’s own product pages, disclosure documents, published rate pages, official FAQs and eligibility documentation.
Third-party publishers may be used to discover that a product exists or to find where a lender documents it. They do not become the evidence for a lender requirement. If a lender does not substantiate a figure itself, we record it as not published rather than repeating what someone else has written about it.
For general educational material, authoritative public sources are preferred, such as the Consumer Financial Protection Bureau.
How the research is reviewed
The process runs in one direction:
Primary sources → AI-assisted research → structured evidence → editorial review → publication
Each record keeps the source it came from, the type of source, the date it was read, the exact wording relied on, and whether each figure is something the lender published or something we worked out. A human reviewer can then check the evidence rather than repeat the research.
Until a human has reviewed a particular version, the site attributes the research to Chris and states editorial oversight, and does not claim the material was reviewed. That distinction is generated by the review system itself, not typed into the page.
Limitations of automated research
Worth being plain about, because these are real:
- Misreading. Automated extraction can misread a page, particularly where a figure is hedged, conditional, or stated in a footnote rather than in the main copy.
- Missed fine print. Substantive terms often live in disclaimers. Research that reads only the marketing copy produces a record that looks confident and is incomplete.
- Context. A rate may be conditional on things the page states elsewhere, such as automatic payments, a minimum initial draw or a location. Those conditions matter and are easy to separate from the number by accident.
- Change. Lender pages change without notice. A record is accurate as of the date it was read, and that date is published alongside it for exactly that reason.
- Judgement. Deciding whether hedged wording on a lender’s own page counts as a product term is a judgement, and judgement is where automation is weakest.
This is why human review exists in the process, and why we publish the source and the date rather than asking you to take the figure on trust.
Why primary sources
Financial product information degrades as it is copied. A figure appears on a comparison site, gets repeated by three others, and within a year it is everywhere and traceable to nobody, still being quoted after the lender has changed it.
Reading the lender’s own documents is slower and produces more blanks. It also produces a record that can be checked, corrected and dated. On a subject where being wrong costs a reader real money, that trade is worth making.