Skip to main content
Alpha Cyanea is in public alpha. We're building in the open — expect rough edges and rapid iteration. See what's live →
Public Alpha

The system of record for computational biology

Every analysis versioned, reproducible, and provenance-tracked — so your results stay trustworthy and re-runnable, across people and institutions. Built on 18 open-source Rust crates and 3,700+ tests. See what's live today.

Reproducible by construction

Every run captures its inputs, code version, runtime, and outputs as a content-hashed record — so any result re-runs with one click and diffs cleanly against the last. No dead environments, no "works on my machine."

  • Hashed inputs with pinned code and runtime
  • One-click re-run, in the browser or on the server
  • Run-to-run diffs you can review

Data sovereignty & federation

Run your own node and keep sensitive data on your infrastructure. Share the analysis — not the data — with collaborators and the wider network, so consortia can work together without moving what they can't move.

  • Self-host with no artificial limits
  • Signed-manifest federation across institutions
  • Cross-node references, citation, and discovery

Team governance & review

Turn shared work into a system your team can standardize on — private spaces, granular roles, review and approval on analyses, and a full audit trail of who ran, changed, and approved what.

  • Private spaces with granular roles
  • Review and approval workflows on analyses
  • Exportable audit logs

Runs anywhere, on real data

One native Rust engine, two execution paths: instant in-browser WebAssembly for interactive work, and server-side compute for datasets that don't fit in a tab. Same code, no reimplementation.

  • In-browser WASM — no setup, no uploads
  • Server-side execution for scale
  • 18 open-source crates, 3,700+ tests

Open & interoperable

Cyanea sits on top of the tools you already use. Import Jupyter notebooks, drive everything from a REST API and CLI, and bring results in from existing pipelines — no walled garden, no lock-in.

  • REST API and CLI for automation
  • Jupyter import; webhooks for events
  • Apache-2.0 licensed, self-hostable

Provenance, lineage & versioning

Every dataset, notebook, and protocol is immutable and content-addressed. Trace any result back through its inputs and transforms, and cite an exact snapshot that will never silently change.

  • Content-addressed, deduplicated storage
  • Inputs → transforms → outputs lineage
  • Citable, immutable version snapshots