Spinorio Lab

A quantum-tuned development environment built on JupyterLab.

Spinorio Lab brings quantum development, AI-powered compilation, and automatic workflow tracking into one JupyterLab workspace. Build circuits with supported quantum SDKs, optimize them for your target hardware, and run experiments on simulators or quantum devices. Inspect circuit states, compare structural metrics, and trace results back through the compilation and execution steps that produced them.

Request Early Access

JupyterLab, ready for quantum development

Build circuits and run experiments on simulators and quantum hardware from your notebook.

Spinorio Lab brings supported quantum SDKs, provider connections, and execution tools into a managed JupyterLab workspace. Write and run Python code with workflow tracking, circuit inspection, and AI-powered compilation built in.

AI-powered compilation, built into your workflow

Optimize quantum circuits for your target hardware through the Lab interface or Python.

Spinorio Compiler uses trained reinforcement learning policies to select and sequence compilation steps for your circuit and target device. Choose your hardware and optimization objective in the Lab, or compile directly from your notebook with the Spinorio SDK.

Your quantum workflow, captured automatically

Track circuit changes, compare structural metrics, and trace each result back to the circuit that produced it.

Spinorio automatically captures your live quantum workflow, from logical circuits and transpilation to job submissions and results, preserving compilation lineage and providing structural metrics to inspect and compare circuit states.

Spinorio Compiler

A quantum compiler built on learned compilation intelligence.

Spinorio trains device-specific policies to determine effective compilation trajectories for quantum hardware. The compiler combines learned policy inference with target-aware evaluation and a single service that can be used from Spinorio Lab or called directly from Python and API workflows.

Learned compilation policies

Device-specific policies learn how to compose compiler transformations for a target processor.

For each supported target and optimization objective, Spinorio trains a reinforcement-learning policy to choose effective sequences of synthesis, mapping, routing, and optimization actions. The learned policy operates above established compiler primitives: it learns how to compose transformations for the current circuit state rather than replacing the underlying algorithms.

Spinorio uses versioned training corpora spanning different algorithm families, circuit scales, interaction patterns, and two-qubit densities. Training infrastructure records model configuration, action-space and observation-space definitions, seeds, dependencies, checkpoints, and evaluation results so policies can be reproduced and compared before production use.

Spinorio maintains policies against explicit target definitions rather than treating hardware as interchangeable. Automatic backend selection can operate over the subset of supported devices available through the user’s provider instance, while fixed-backend compilation remains available when the target is already known.

Candidate circuits can be compared by target validity, native operations, connectivity, two-qubit operations, routing overhead, depth, estimated duration, and estimated execution quality. Hardware and calibration context can be attached to the experiment so the recommendation remains tied to the conditions under which it was produced.

Spinorio Lab provides a clean interface for selecting a circuit, optimization objective, and target hardware, then inspecting recommended candidates and compilation evidence. Programmatic workflows call the same compiler service through the Spinorio API and Python SDK—for example with spinorio.qcompile().

Spinorio Workflow

A complete record of the quantum experiment.

Spinorio connects logical circuits, compilation states, hardware jobs, and results into one inspectable research record. Structural metrics and execution context stay attached to the exact circuit state that produced them, so transformations and outcomes remain traceable across the experiment lifecycle.

Logical circuit

Immutable source states, circuit structure, metrics, and experiment context.

Spinorio preserves each circuit version as an immutable state and computes a structural profile for it. Alongside circuit scale and depth, the profile records program communication, critical depth, entanglement ratio, parallelism, liveness, and measurement behavior. These descriptors remain attached to the exact circuit state, making structural changes across compilation and execution directly traceable.

Compiled circuit states remain linked to their source state so researchers can inspect how structure, mappings, depth, two-qubit operations, and other metrics changed. When Spinorio Compiler is used, model version, objective, target definition, action sequence, candidate metrics, and recommendation evidence can be preserved with the compilation experiment.

Hardware execution is connected to the exact circuit state that was submitted. Spinorio can preserve the backend, provider job identifier, execution status, timestamps, runtime metadata, and available target or calibration context without collapsing those details into a separate disconnected job history.

Results are stored as the final stage of the same experiment lineage. Researchers can move from an observed result back through the hardware job and compiled circuit to the original logical state, keeping the evidence needed to inspect, compare, and reproduce the experiment.