Privacy, in plain language.

1. Who operates SourceArc

Prajana AI operates SourceArc. During the controlled beta, privacy and account requests are handled directly at hello@prajana.ai.

2. Information we handle

Account information

We receive the verified email address, phone number, or basic Google account identity you use to sign in. We also retain your account status.

Content you choose to provide

This includes pasted notes, public website links and extracted text, uploaded documents, campaign settings, generated drafts, edits, versions, quality feedback, and publishing approvals.

Provider and publishing connections

If you bring an API key or connect a publishing account, we store the credential in encrypted form and display only masked identifying details. We do not expose the secret again after entry.

Operational information

We record content-free events such as model choice, token counts, latency, error category, quota use, and feature actions. Logs are designed not to contain source text, generated text, or credentials.

3. How we use information

• Authenticate you and enforce controlled-beta access.
• Extract sources and generate, edit, save, and publish drafts.
• Show history, versions, usage, cost estimates, and reliability.
• Protect the service, enforce limits, and investigate failures.
• Improve prompts and workflows using content-free measurements and explicit feedback.

We do not sell personal information. We do not use consumer campaign content to train public models.

4. Model providers and subprocessors

Generation sends the source material needed for your request to the model provider you select. Current options may include Google, OpenAI, Anthropic, and OpenRouter. When you use your own provider key, that provider relationship is also governed by your account terms with them. Provider processing and retention practices can differ, so avoid submitting secrets or unnecessary sensitive personal data.

SourceArc runs on Google Cloud and uses Firebase Authentication for sign-in. Publishing features send approved content to the destination you explicitly connect and confirm.

5. Storage, security, and retention

Traffic is protected in transit. Persisted source content, generated output, extracted document text, provider credentials, and publishing tokens are encrypted at rest. Access is scoped to the signed-in owner, with a single administrator for beta operations.

Active campaign history is retained until you delete it. Deleted campaigns stop appearing in the product immediately and are scheduled for removal from primary storage. Temporary processing artifacts are removed automatically. Residual encrypted copies may remain in security backups until the applicable backup cycle expires.

Exact purge and backup windows are being validated before the beta opens beyond verified beta users. Contact us if you need a deletion request tracked to completion.

6. Optional public-site analytics

Public-site analytics are off by default. If you allow them, we use Google Analytics with IP anonymization, advertising signals disabled, and no campaign content or provider credentials. Your choice is saved in your browser and can be changed through “Privacy choices” in the site footer.

7. Your controls

You can delete campaigns, revoke provider credentials, disconnect publishing accounts, decline analytics, and request access, correction, or account deletion by emailing hello@prajana.ai. Beta users must contact the administrator for account-level requests.

8. Changes

We will update this notice as SourceArc adds providers, media, public API access, or commercial plans. Material changes will be dated here and communicated to active beta users when appropriate.