Skip to main content
Introducing The Quantum Infrastructure Cloud Learn more

SynapseX AI

AI agents for scientific discovery

From catalysts to orbits: agents that fetch real data, run the calculation and show you every number they computed, with a person in charge.

Capabilities

One workspace for agents, automations, skills and libraries

Agents

  • Catalogue
  • Create
  • Up to 8
A guide to the capabilities, drawn for this page. Not a screenshot of the app.

Use cases

Science, from the molecule to the orbit

Each card names the SynapseX tool that does the work. The tool fetches or computes the numbers, and the agent explains them.

A glass flow reactor with coiled tubing carrying a bright yellow-green solution in a fume hood

Chemistry

From a molecule name to its quantum circuit

Molecules come from PubChem and crystals from the Crystallography Open Database, drawn in 3D. A molecule can go straight to a quantum circuit plan for its ground-state energy.

Fluorescence microscopy of human cells, with blue nuclei, green filaments and magenta mitochondria

Biology

Proteins, genes and the papers behind them

Structures from the PDB and AlphaFold DB with measured contacts and confidence, records from UniProt, Ensembl, Reactome, ClinVar, GEO and PubMed, and genome tracks in a browser view.

Earth from low orbit at the line between day and night, with a satellite in the foreground and the Moon beyond

Space

Real ephemerides beside your own model

Where a planet, moon or spacecraft was, is or will be, read from JPL Horizons and drawn in 3D at real scale, next to an orbit you integrate yourself.

A rocket engine firing horizontally on a night test stand, its plume lighting clouds of steam

Rocket science

Trajectories with the real atmosphere

Dynamical systems integrated and plotted, with air temperature, pressure and density up to 86 km from the 1976 U.S. Standard Atmosphere.

A faceted crystal refracting violet and teal light beside a lattice model

Modern physics

Fields, wave packets and error-correcting codes

Crystal structures, surfaces and vector fields, Schrödinger wave packets stepped live on a lattice, and surface and repetition codes with their decoding.

A misty rainforest river at dawn with an environmental sensor station on the bank

Natural science

Field data, mapped and analysed

Maps and choropleths, scientific data files such as VTK, coordinate tables and XYZ, and embedding spaces, with Python in an isolated sandbox for the analysis.

AI agents in catalysis research

Hand the routine of computational chemistry to agents

Much of a catalysis study is routine: finding structures, setting up each calculation, launching runs, collecting the numbers and comparing them. Steps a researcher strings together by hand, often over days, an agent sets up in one conversation, and every number it reports was computed.

Close-up of metal nanoparticles on a porous catalyst support, with gas bubbles forming in a clear liquid

Step 1 of 4 · Find

Input
SMILES · XYZ · PDB
Source
PubChem conformer
Checked
Size and format

Ground-state energies are implemented today. Reaction barriers, transition states and surface models are not.

  • Today

    Faster simulation

    Agents take on the routine of a computational study: structures, set-up, runs and the collection of results.

  • Today

    Patterns in large datasets

    Upload data or index your papers in a Library, then let an agent explore it with Python in a sandbox and plot what it finds.

  • Vision

    A lab in the loop

    Agents that propose the next experiment and hand it to lab automation. This is not in the product today.

What SynapseX brings

An AI workspace built around how science checks itself

11
public science sources wired into the chat, from PubChem to JPL Horizons
8
agents that can work on one task at the same time
4
ways in: web chat, Lab, desktop and mobile apps, CLI and API
  • Numbers that were computed, not written

    Distances, atom counts, confidence scores and energies come from code. The model is told never to make one up in prose.

  • Science sources, wired in

    PubChem, the Crystallography Open Database, the PDB, AlphaFold DB, UniProt, Ensembl, Reactome, ClinVar, GEO, PubMed and JPL Horizons, each within its published rate limits.

  • Chemistry to quantum circuits

    A molecule becomes a circuit plan and runs on a free CPU simulator through QCOS. Execution on quantum hardware is switched off.

  • Agents that check each other

    A task can fan out to parallel agents, combine their results and vote on them in a verify stage that fails closed.

  • Experiments you can retrace

    Experiment plans, versioned datasets, run comparison and evidence receipts in SynapseX Lab.

  • Work that keeps running

    Scheduled tasks with a tool allowlist and a time limit per run, suspended after repeated failures.

How an agent works

Real tools, with a person in charge

  • Credentials stay out of the model

    Agents name a connector and an operation. Tokens stay behind a gateway, and secrets are redacted.

  • Outside content is treated as data

    Tool output is framed as untrusted, and probing is audited and refused. This is layered defence, not formal taint tracking.

  • Consequential actions wait for you

    Actions are classed as read, write, publish, financial or destructive, and publish and financial actions need approval.

Model programme

Research models, in development

SoftQuantus is specialising models for quantum and scientific work with SFT and DPO. Until a programme model passes its evaluation, SynapseX runs models chosen for each kind of work.

  • Training on European HPC

    SoftQuantus holds a EuroHPC AI Factory allocation (EHPC-AIF-2025LS13-019) for training compute.

  • A gate before anyone uses it

    A programme model serves only after it passes its evaluation. None has passed yet, so none serves today.

  • Size, when it serves

    We will publish the size and results of the largest programme model when it serves, not before.

Method papers: DPO and RAG.

Vision

What we are working towards

None of these exists in the product today.

  • Vision

    A lab in the loop

    Agents that plan the next experiment from the last result and hand it to lab automation.

  • Vision

    Dedicated model deployments

    A model running only for your organisation, inside its own boundary.

  • Vision

    Confidential computing

    Agent machines in hardware enclaves, with keys you hold.

  • Vision

    A standing verifier

    Observers beside every turn that check the work, recall memory and track the goal.

  • Vision

    Tainted-egress tracking

    Once a process reads untrusted content, its network access waits for a person.

Trust

What we state, and what we do not hold

  • Email and password or Google sign-in. SAML and SCIM on request.
  • 99.9% availability target. This is not a contractual SLA.
  • Logs kept for 30 days. A DPA is available on request.
  • We do not hold SOC 2 or ISO 27001.
Request enterprise access

Sign up for SynapseX updates

Film and photographs made for this page by SoftQuantus. They illustrate each field and do not show SoftQuantus hardware, staff, customers or results.