Untilum seals any file until a chosen future date so that nobody can open it early — not the platform, not the storage provider, not even the author. This is guaranteed by construction, not by company policy. Everyone sells a promise; we sell mathematics that keeps it.
Every existing "letter to the future" is built on trust in an intermediary. Ordinary encryption answers "who can read this" — never "when". That category doesn't exist on the market yet.
Timelock encryption became practical with the launch of drand quicknet (2023); pay-once permanent storage matured in parallel. Building this product became possible only within the last two years. The category is empty and the window is open.
Carrot · Munich, Germany.
Product, protocol governance, fundraising.
Crypto core, threat model, internal pentesting.
TypeScript/Rust SDKs, web/desktop/CLI, cross-implementation CI.
What's done, and what's a milestone target, not a current fact.
10 RFCs + 3 wire specs in public Last Call, closing 2026-08-10
Two independent implementations (TypeScript, Rust) — byte-identical containers enforced in CI
4 critical crypto tests green; internal penetration test and critical review completed (July 2026)
External security audit — planned before public launch, not yet started
Public launch with paid sealing and 2–3 B2B pilots — milestone target, not current state
Convertible loan with SAFE economics (cap + discount), converting into equity at the next qualified financing. Indicative use of funds: team of 3 FTE on milestone-based compensation (~48%), external security audit (~12%), public launch (~10%), protocol publication (~4%), legal (~4%), downside reserve (~22%).
Milestones: independent audit report, public launch with paid sealing and 2–3 B2B pilots, protocol Last Call, operational break-even around month 16–17.
✓ Request sent — check your inbox to confirm data room access.