Privacy Policy

Last updated: July 10, 2026 (v2)

KnowledgeRetention is operated by ChAImp, a nonprofit organization, 720 Seneca St, Ste 107 #964, Seattle, WA 98101.

What we collect

• Account details: name, email, hashed credentials, two-factor enrollment.
• The knowledge your organization chooses to capture: interviews, voice notes, uploaded files, forwarded emails, and the structured items derived from them.
• Waitlist and support submissions you send us.
• Operational logs needed to run the service securely (see the audit log below).

Please don't submit sensitive personal data— Social Security numbers, government IDs, full financial account numbers, passwords, or health records. The service doesn't need them, actively prompts against them, and redacts them where detected. Anything submitted anyway is processed as customer content at your organization's direction and responsibility (see the Terms, “Sensitive personal data”).

Our commitments

• We never sell or rent your data.
• We never use your data — even anonymized or aggregated — to train any AI model. Our AI providers process it under terms that prohibit training on it.
• One organization can never see another organization's data — isolation is enforced at the database layer (row-level security) on every table.
• Anonymized, non-identifying analytics used to improve the service never involve model training.

Where your data lives

Third-party processors we use to deliver the service:

• Neon (AWS) — database
• Vercel — hosting, file storage, DNS
• Anthropic — AI structuring and answering (no-training terms)
• AssemblyAI — voice-to-text transcription
• Resend — transactional email
• Stripe — billing (paid plans; we never see full card numbers)
• Cloudflare — abuse protection on public forms

We don't control these providers and may add or replace providers over time; material changes will be reflected here.

AI providers in detail

Voice-to-text — AssemblyAI

Voice notes are transcribed by AssemblyAI. At the service level, ChAImp has:

• Opted out of AssemblyAI's Model Improvement Program — your audio and transcripts are never used to improve their models. Learn more
• A 1-day time-to-live (TTL) on audio recordings— audio is deleted from AssemblyAI's systems within one day. Learn more
• A signed Business Associate Agreement (BAA) with AssemblyAI. Learn more

AssemblyAI's own policies:

• Privacy Policy
• Data Processing Addendum
• Data retention & model training
• Trust & compliance center

All other AI — Anthropic

Knowledge structuring, answering, and continuity briefs are powered by Anthropic under terms that prohibit training on your content. Their policy: Anthropic Privacy Policy.

Security

• Two-factor authentication is mandatory on every account, on every plan.
• Deny-by-default row-level security scopes every record to your organization.
• An immutable activity log records administrative and knowledge changes — who, what, when — visible to your organization's owners and retained to support troubleshooting and the investigation of improper activity.
• If we confirm a security breach involving your data, we will notify you without undue delay and take reasonable steps to mitigate it.

Retention, export & deletion

Export your organization's data any time (one click, every plan). Canceled accounts retain data for 90 days for export, then it is deleted. You are responsible for keeping your own copies of data critical to your operations. Deletion requests: email us below.

Contact

privacy@knowledgeretention.help
ChAImp · 720 Seneca St, Ste 107 #964, Seattle, WA 98101