Know what is collected before you forecast.

Plain language summary

ATAR Forecast is operated by Kanishk Joshi in NSW, Australia. It stores the information needed to create, reproduce and improve a forecast. It also stores contact details and a compact academic profile that ATAR Forecast may use to contact you about relevant HSC tutoring opportunities when the required permission is given. We do not ask for a full name, exact date of birth, street address or school login.

1. What we collect

Account and contact details

Clerk supplies a stable account identifier and account email when you sign in. If you choose Google sign in, Google supplies the account details needed by Clerk. ATAR Forecast does not receive or store your password. The verified account email is used as the forecast contact method, so the form does not ask you to type it again. A phone number and 4 digit postcode are required. Do not enter a street address. The forecast record does not request a full name.

Academic and forecast information

We store your target HSC year, school, courses, ranks, cohort sizes, calculated relative rank positions, estimated marks, five annual ATAR scenarios, central ATAR and range, ATAR goal, estimated gap, current preparation and potential improvement subjects.

Audit and future calibration information

Every new forecast is a separate historical record. We store its creation time, model version, data schema version, privacy and terms versions, consent time and consent purposes. If you later provide actual HSC marks or an ATAR, those fields may be attached for accuracy testing. Old forecasts are not silently replaced.

Optional forecast feedback

After viewing a result, you can rate how closely it matched your expectation. If you give a low rating, we also ask for the ATAR you expected, where that expectation came from and optional comments. This feedback is linked to that forecast and used to identify where the beta model may need improvement.

HSC alumni calibration study

The account free study stores a random submission identifier, school, HSC year, actual and predicted ATAR, courses, ranks, cohort sizes, HSC marks, optional trial marks, survey answers, optional feedback, model version, consent version and timestamp.

2. Why we use it

We use records to deliver and reproduce forecasts, calculate a goal comparison, suggest one or two possible study priorities, improve and calibrate the model, keep accounts separate, respond to privacy requests, analyse optional forecast feedback, prevent misuse and maintain the service.

3. Tutoring contact and service providers

ATAR Forecast may use a permitted student profile containing the account email, phone number, postcode, HSC year, school, subjects, ranks, cohort sizes, forecast marks, current ATAR range, ATAR goal, estimated gap, preparation status and identified improvement opportunities to contact the student about relevant tutoring. This permission is a condition of receiving the free forecast and is stated in the single checkbox shown before rank information is entered.

Clerk processes account and authentication data. Google processes data when you choose Google sign in. OpenAI Sites and Cloudflare provide application hosting, database and delivery infrastructure. Their systems may process data outside Australia, including in the United States and other service regions. We do not publish student profiles or share them with schools, UAC, NESA or a third party tutoring provider unless an additional disclosure and permission process is introduced.

4. How records are protected

Forecast routes require platform authentication. Administrative views and exports require a separate server side allowlist. Tutoring lead exports exclude records without the required contact permission. Input validation, database relationships, timestamps and versioned consent events keep records traceable.

No online service can promise absolute security. Suspected access incidents are investigated and assessed under applicable breach notification requirements.

5. Retention

Identifiable forecast and lead records are reviewed after the student’s target HSC year and are intended to be deleted or deidentified within 18 months after that year unless they remain necessary for a request, dispute or legal obligation. Deidentified calibration information may be retained for longer while it is useful for measuring model accuracy. A verified deletion request can shorten this period.

6. Access, correction, deletion and withdrawal

You can ask what information is linked to you, request a correction, ask for deletion, withdraw future tutoring contact or raise a complaint. The unsubscribe page does not require an account, password or reason. It immediately suppresses matching forecast records from future tutoring lead use. Withdrawal does not affect processing already completed. Identity may need to be verified before linked records are disclosed or changed.

7. Students under 18

This product requires users to be 15 or older. A student should only agree if they understand what information will be collected, why it is used and how to withdraw. Acceptance of the terms, privacy policy and tutoring contact is combined in one affirmative checkbox and is required to access the free forecast. A student who does not understand the choice should ask a parent or carer to review it with them before continuing.

8. Changes and contact

Material changes receive a new version date. Each submission keeps the privacy and terms versions shown when it was created. Contact Kanishk Joshi at kanishk.joshi2702@gmail.com or use the privacy request form for access, correction, deletion, withdrawal or complaints. We will investigate and respond within a reasonable time.