Carbon

    Platform

    Data Layer

    Versioned · Governed · Queryable

    Versioned, governed research data. From instrument readouts to analysis-ready datasets in one pipeline, query, version, and share without copy/paste labs.

    A result is only as trustworthy as your ability to say where it came from. Carbon's data layer captures every readout as a versioned dataset that remembers the instrument, run, and configuration that produced it, so reproducing a measurement is a query, not an archaeology project.

    Every dataset
    Versioned
    Every dataset
    Run lineage
    Full
    Run lineage
    Source of truth
    One
    Source of truth

    Carbon DB

    Design your data model. No SQL required.

    Every team gets a managed Postgres, locked down and easy to use. Lay out tables and relationships visually, and store things plain Postgres can't, like images, files, and waveforms, as first-class types. Drag the tables around.

    Loading schema…

    One pipeline from readout to result.

    Instrument output lands as versioned datasets, immediately queryable from analysis and applications.

    Lineage you can audit.

    Every dataset knows which instrument, run, and configuration produced it.

    Share without copies.

    Grant access instead of emailing files, one source of truth across the team.

    INSTRUMENTCONFIG · #A3F2ACL + AUDITDATASETRUN 0142 · V4LINEAGE INTACTPYTHONDASHBOARDEXPORT / APIPRODUCED BYEVERY VERSION KNOWS ITS INSTRUMENT, RUN, AND CONFIGURATIONQUERY, NOT FOLDER ARCHAEOLOGY

    How it works

    01

    Capture

    Every readout lands as a versioned dataset, tagged with the instrument, run, and configuration behind it.

    02

    Query

    Find data by lineage, not folder archaeology. Analysis and applications read the same source of truth.

    03

    Govern

    Grant access instead of emailing files. Every read and write is recorded against a name.