Privacy
Last updated May 24, 2026
What we collect
- Account: name, email, organisation name. Required to sign in.
- Company data: anything you put into a kigya project — URLs, idea briefs, voice interview audio, founder-voice samples, docs, messages.
- Operational metadata: agent decisions, tool calls, audit events, spend per agent.
- Connected accounts (only what you authorise): Stripe (read-only metrics), Slack (workspace bot token), ad accounts if you opt in.
How we use it
To run your AI co-founder. Your data is fed to LLM providers (OpenAI by default, optionally Anthropic via the Claude Agent SDK, or any BYOA model you connect) only to fulfil tasks you've queued or cycles you've enabled.
What we don't do
- We don't train models on your data. Ever.
- We don't sell or share your company data with third parties beyond the LLM providers you've enabled.
- We don't post to your external channels (Twitter, email, ads) without an approval you signed off on.
Where it lives
Default deployment is US-east on Supabase + Vercel. Corporate Project mode and Enterprise plans route to EU-west-1 with signed BAAs available.
Your rights
Export everything anytime (Settings → Export). Delete everything by emailing privacy@kigya.com — we honour deletes within 7 days.