Dzamisa Pilot

Privacy Notice

Version 1.0 · Effective date: September 10, 2026 · AI-processing disclosure ai-1.0.

This notice explains what information Dzamisa collects from participants in the Dzamisa Pilot, why it is collected, who can see it, and how you can ask for it to be corrected or deleted. It applies to the pilot application page, the access-code page and the decision review page. It does not describe any other activity.

1. Who is responsible for your information

Dzamisa is a service operated by CDiGlobal. CDiGlobal operates the Dzamisa service and decides how and why the information described here is used. Questions, requests and complaints can be sent to reviews@dzamisa.ai.

Whether a full legal address is legally required in this notice, and which business address should be used if it is, is unresolved—not verified. No personal or home address is displayed.

2. What we collect

Through the pilot application page (/volunteer): your full name, professional email address, organization, job title or role, country or region, industry or sector, an optional LinkedIn profile address, your written reason for wanting to take part, and a record of the acknowledgements you tick — that you can provide feedback, that you agree to the Pilot Participation Terms and this Privacy Notice (including the version shown to you and the moment you accepted), and that you will not submit sensitive information.

Through the access page (/access): the access code you enter. Codes are checked against the invitation record, which holds the code, the email address it was issued to, its expiry, and a count of how many times it has been used. A validated code is held in your browser for the rest of that browsing session so that you are not asked for it twice; it is cleared when you close the browser.

Through the review page (/review): your name, email address, optional organization, the access code used, and everything you write about the decision — its title, background and context, options considered, what is at stake, timing, any AI involvement, known risks, constraints and your desired outcome — together with your acknowledgements and the date and time of submission.

Technical information: our hosting, database and sign-in providers may process technical request, authentication and security-log information as part of operating the service. This may include network address (such as an IP address), request time, browser and device information where available, and the page or endpoint requested. This information is generated and retained by those providers under their own schedules; we do not use it to build a profile of you.

We ask you not to provide confidential, regulated, highly sensitive or unnecessary personal information in a decision scenario. Please describe your decision in a way that does not identify other people or disclose material you are not free to share. Note that the system does hold personal information about you separately — your name and email address, taken from your approved application — so that we can return the review to you.

3. Why we use it

4. AI-assisted processing (disclosure version ai-1.0)

The content of a decision scenario you submit may be processed through the external Lovable AI service, solely to create an editable draft for a human reviewer. This happens only where you have given consent by ticking the AI-processing consent box on the review page immediately before submitting. Consent is recorded together with the moment you gave it and the version of this disclosure you were shown.

AI-generated content may be inaccurate or incomplete. It is advisory only. Every AI-assisted output is evaluated, edited and approved by a person before it forms part of a review, and no final decision is made automatically. The accountability for your decision remains with you and your organization.

Your name, your email address, your organization, your access code, consent metadata and the administrator's identity are excluded from the fields sent to the AI service. Only the decision-submission text you write is included. Free-text may still contain names or other identifying details if you put them there, because the current implementation does not automatically remove personal information from free-text. Please do not put confidential, regulated, highly sensitive or unnecessary personal information — yours or anyone else's — into that text. We separately retain your name and email address from your approved application so that we can return the review to you.

Decision scenarios submitted before this disclosure took effect are marked as pre-consent records and are technically blocked from being sent to any AI processing. They will only be processed with AI if fresh consent is obtained from the participant.

The current AI request uses the OpenAI-family model identifier openai/gpt-6-astra through Lovable AI and is configured with store: false. That application setting asks the AI service not to store the generated response for later retrieval. It does not mean that all temporary security or abuse-monitoring logs are prevented. The exact retention periods for any such provider logs are unresolved—not verified.

Lovable states that information submitted by an app's end users is not used to train Lovable's AI models and that its agreements restrict third-party model providers from using transmitted content for training. The Dzamisa implementation additionally sends store: false. Dzamisa will not authorize submitted content for model training without separate, explicit opt-in consent. No such training consent is currently requested.

The administrator-only generation record stores the original generated draft, the case reference, administrator account reference, time, model, ruleset version and outcome. It does not store participant identity fields in that generation record. Its automatic deletion period is unresolved—not configured.

5. Who can see your information, and who processes it for us

We do not sell your information and we do not share it for advertising. We may disclose it where we are legally required to, or to establish or defend legal claims.

6. Cookies, analytics and technical logs

This site uses no advertising cookies and no analytics or tracking tools. Essential authentication and session technology may be used where necessary to make the service work: the pilot administration sign-in stores a session in your browser so that an administrator stays signed in, and the access page stores your validated invitation for the current browsing session only.

This notice will be updated before any analytics, advertising or tracking technology is added.

Our hosting, database and sign-in providers keep their own technical request, authentication and security logs, which typically include information such as network address, time and page or endpoint requested.

7. Where information is stored

The connected project configuration identifies the pilot database and sign-in service as a Lovable-managed Supabase project. Its database connection host is aws-0-us-west-2.pooler.supabase.com, which verifies the configured database-hosting region as AWS us-west-2 (US West, Oregon), United States. The site itself is served through Lovable's hosting on Cloudflare's global network, so site requests may be handled at a location near you even though the database connection configuration identifies us-west-2.

8. How long we keep it

Authentication and security logs may be generated by Supabase for database and sign-in activity, Lovable for managed infrastructure and AI-gateway activity, and Cloudflare for hosting and request security. Their exact retention periods are unresolved—not verified from the connected configuration and project documentation available to us. These provider logs are separate from the retention periods above, and we do not use them to build a profile of you.

9. Security

We take reasonable steps to protect this information. Reading, editing and administering stored pilot records is restricted to authorized administrators. Public users may submit information only through the designated application and decision-review forms. They cannot browse, retrieve, edit or administer stored records. Invitation codes expire and can be withdrawn.

No online service can be guaranteed to be secure, and we do not make guarantees, certify compliance with any standard, or claim any particular certification for this pilot. Please keep sensitive material out of your submissions.

10. Your requests

You can ask us for a copy of the information we hold about you, ask us to correct it, ask us to delete your application or your submitted scenario, or withdraw from the pilot. Write to reviews@dzamisa.ai from the address you applied with, or tell us which address to check. We aim to acknowledge verified privacy requests promptly and respond within 30 days. If more time is reasonably required, we will explain the reason and expected response date. We may keep a minimal record that a request was made and met. Depending on where you live, you may also have the right to complain to a data protection authority.

11. Adults only

The Dzamisa Pilot is for adults. Applicants and participants must be at least 18 years old. We do not knowingly collect information from anyone under 18; if we learn that we have, we will delete it.

12. Changes to this notice

If we make a material change — for example to what we collect, why we use it, who we share it with, or anything to do with AI processing — we will publish an updated version with a new version number and effective date on this page, and email pilot participants at the address they applied with. If a material change introduces a new purpose, external AI processing, model training, or materially different disclosure of information, we will request renewed affirmative consent before applying that change to previously submitted identifiable information where required.

13. Contact

reviews@dzamisa.ai

This notice describes the current data practices of the Dzamisa Pilot. It does not represent a compliance certification.