THE DATA CONSORTIUM

Simulation is just a guess.

Ground your internal AI models in thousands of physical, standardized data points. We provide the highest-fidelity, machine-readable training data for materials science, mapping both breakthroughs and invaluable synthesis failures.

Read the API Docs View Data Schema

consortium_firehose_stream.py

[SYSTEM] Connecting to stream: wss://stream.solbion.com/v1/telemetry...

[SYSTEM] Stream established. Streaming live synthesis records.

[RECORD] id: run_9024 | type: spark_ablation | status: FAILURE_OVERHEATED

[RECORD] id: run_9025 | type: magnetron_sputter | status: SUCCESS_NOMINAL

BUFFER: NOMINAL 24.2 EPS (EVENTS/SEC)

GROUND TRUTH

Built for Machine Learning

Structurally Unified

No more parsing fragmented PDFs or messy ELNs. Every run across our physical automated labs is logged into a strictly typed schema, ready for immediate ingestion into your data lakes.

The Negative Space

Algorithms learn faster from failure. We do not cherry-pick successes. Our dataset includes comprehensive logs of failed syntheses, anomalous degradation, and out-of-bounds reactions.

High-Fidelity Telemetry

We log far more than just the end material. Access microsecond-level telemetry, including vacuum pressures, precursor flow rates, plasma energies, and real-time metrology feedback.

SAMPLE PAYLOAD

Inspect the Telemetry

Every API response delivers the complete lifecycle of a physical synthesis run, linking hardware parameters directly to structural metrology.

{
  "experiment_id": "run_7894_bimetallic_cu_ni",
  "timestamp": "2026-07-06T14:32:01Z",
  "outcome_flag": "FAILURE_NUCLEATION_ARREST",
  "synthesis_parameters": {
    "technique": "spark_ablation",
    "carrier_gas": "Ar_99.999%",
    "flow_rate_sccm": 150.0,
    "spark_energy_joules": 0.45,
    "chamber_pressure_mbar": 1.2e-3
  },
  "inline_metrology": {
    "smps_particle_size_distribution": {
      "median_diameter_nm": 2.4,
      "geometric_standard_deviation": 1.85
    },
    "optical_emission_spectra": {
      "peak_wavelengths_nm": [324.7, 327.4, 341.4],
      "intensity_au": [4502, 3811, 890]
    }
  },
  "metadata": {
    "anomalies_detected": ["electrode_wear_variance"],
    "machine_id": "node_amsterdam_02"
  }
}

ACCESS

Join the Consortium

Academic & Seed

Universities and Pre-Series A Startups

STARTUP/ACADEMIC

Access to foundational, non-proprietary datasets
Rate-limited REST API access
Monthly batch CSV/JSON exports

Apply for Access

Enterprise R&D

Fortune 500 Chemical & Semiconductor Firms

ENTERPRISE

Full firehose access to the negative space database
Real-time WebSocket streaming from public lab nodes
Dedicated VPC peering and integration engineering
IP indemnification

Contact Sales

INITIATE API ACCESS

Stop modeling in a vacuum.

Your active learning algorithms are only as good as the physical data feeding them. Connect to the Solbion database and start mapping reality.

Generate API Key