KlyrOpen workspaceYour customer research, handled the way you’d want.
Klyr only works if you’re comfortable putting real customer interviews into it. Here’s exactly what happens to that data, in plain language, no legalese.
Never used to train models
Your transcripts, notes, and generated artifacts are never used to train AI models, not ours, and not the providers’. Klyr runs on the AI providers’ API tier, which excludes your inputs and outputs from model training.
Isolated to your workspace
Research is scoped to the workspace that uploaded it and enforced at the database layer with row-level security. One workspace can never read another’s data.
Yours to export or delete
You can request a full export or permanent deletion of your account and workspace data at any time. Deletion removes the underlying files, not just the references.
Private by default
Reports are visible only to your workspace. A report becomes shareable only when you explicitly create a share link, and shared reports are excluded from search engines.
Built on trusted infrastructure
Klyr runs on infrastructure providers (Vercel and Supabase) that maintain their own SOC 2 Type II reports, with data encrypted in transit and at rest. We don’t sell your data, ever.
Citations, not guesses
Every claim in a synthesis links to the exact source quote it came from, and any theme without at least two verified citations is dropped, so nothing in your roadmap is a model fabrication.
Your data is never used to train AI models, it’s isolated to your workspace, you can export or delete it anytime, and reports stay private until you choose to share them.