DHULFIKAR STEWARD — Trust, Evidence & Current Status
New to STEWARD? Start here · How it works · Glossary · Test your understanding · What it is and isn't · Frequently asked questions
Why should anyone trust STEWARD?
Not simply because DHULFIKAR says STEWARD is intelligent. Trust needs to be earned through verifiable methods, clear boundaries and evidence.
What we intend to enforce
- Fact ≠ assumption. The two must be kept separate.
- Model ≠ truth. A score or AI result remains a tool.
- Internal review ≠ external approval. A partnership, regulatory approval or professional authorisation must never be invented.
- Product interests ≠ customer interests. Revenues and conflicts of interest must become visible.
- Design ≠ proven operation. The existence of a feature does not automatically prove its operational effectiveness.
- Human escalation. Specialist questions must be able to reach authorised people.
- Recovery. Ask not only “what if it works?”, but also “what if it fails?”.
Current status
The STEWARD core architecture, capability register and several fail-closed contracts and tests are documented internally. Public pages provide pre-launch information. Not every commercial or regulated feature is operational, and having a page on the website does not constitute a licence, partner acceptance or production approval.
What still needs to be demonstrated?
- Real user testing of understanding.
- Operational outcomes measurement.
- Privacy and data-scope implementation for each product.
- Specialist routing and assignment of responsibilities.
- External partner and regulatory requirements where applicable.
- Up-to-date product status for each module.
Status: STEWARD is currently a pre-launch decision-support and evidence architecture. It does not replace a banking licence, lawyer, accountant, investment adviser, lender, insurer or other authorised professional.
How do we prevent old evidence of understanding from remaining valid after the text changes?
STEWARD's design binds comprehension-test evidence to the exact content version that was tested. Material changes can therefore trigger a mandatory retest. A misunderstanding can only be treated as resolved once new test evidence is linked to the corrected content version.
Read about version-linked evidence of understanding and retesting →
Price of this information page
€0. This information page is public and free. Paid STEWARD services have their own price or pricing status on their respective product pages; this free explanation does not change those prices.
What this information page does not do
- This page provides information; it does not grant any licence, approval or financial authority.
- It does not guarantee that every visitor automatically understands the content; actual understanding needs separate testing.
- It does not replace an authorised professional where law, contract, risk or complexity requires human expertise.
- It does not turn a pre-launch or internally tested capability into a live, production-ready service.
Last structural STEWARD content audit
7 October 2026: At that time, 51 of 51 current STEWARD pages were publicly checked against seven basic criteria: route to understanding, price or pricing status, current status, explicit boundaries, next step, FAQ link and glossary link. All seven checks covered 51/51 pages.
This coverage did not establish actual user understanding. In that earlier repair round, 46 page bodies were changed. The previous hash baseline was consequently treated as stale, and a new post-governance baseline was created and checked at 51/51 exact matches. Genuine comprehension retesting remains necessary for amended information.
Translation note: this historical audit statement refers to 7 October 2026. Subsequent language revisions require their own new content-version baselines; the historical 51/51 result is not a certification of this translation.