SoftQuantus is a European industrial research and development group working at the intersection of quantum computing, artificial intelligence, high-performance computing and scientific software systems.
The group develops methods, software infrastructure and experimental workflows for executing, optimizing, benchmarking and validating advanced computational workloads across quantum processors, quantum simulators, GPUs, HPC systems and cloud environments.
RESEARCH AREAS
Quantum Computing and Quantum Software
Our quantum computing research focuses on the complete lifecycle of quantum and hybrid quantum-classical workloads, including:
• quantum circuit design, compilation and optimization;
• backend characterization and hardware-aware execution;
• quantum algorithm benchmarking;
• noise modelling and execution stability analysis;
• error mitigation and result validation;
• hybrid quantum-classical algorithms;
• quantum simulation using CPU, GPU and HPC resources;
• multi-provider quantum workload orchestration;
• reproducibility and verification of quantum experiments;
• resource estimation, execution cost analysis and performance comparison.
QCOS
SoftQuantus develops QCOS, a provider-independent orchestration and execution infrastructure for quantum computing.
QCOS is designed to coordinate workloads across different quantum frameworks, simulators, emulators and hardware providers. Its research objectives include backend-independent execution, workload optimization, benchmark standardization, experimental reproducibility and the generation of verifiable computational evidence.
The platform investigates how quantum workloads can be executed consistently across heterogeneous systems while preserving information about the software environment, circuit transformations, backend properties, execution parameters, measurements and resulting data.
Hybrid HPC–Quantum Computing
The group studies architectures that combine quantum processors with classical computing infrastructure.
Research activities include:
• GPU-accelerated quantum simulation;
• HPC-based circuit simulation and benchmarking;
• hybrid CPU–GPU–QPU execution pipelines;
• distributed scientific workloads;
• classical optimization supporting quantum algorithms;
• scheduling and resource allocation across heterogeneous compute systems;
• integration of quantum execution with cloud and supercomputing environments.
Artificial Intelligence, Deep Learning and Large Language Models
SoftQuantus also develops advanced AI systems based on deep learning, large language models and autonomous agents.
Research topics include:
• training and adaptation of large language models;
• distributed model training and inference;
• model quantization and inference optimization;
• retrieval-augmented generation;
• scientific and technical language models;
• multi-agent and agentic AI architectures;
• LLM orchestration and tool execution;
• integration of AI agents with APIs, databases and scientific software;
• AI-assisted scientific experimentation;
• automated analysis of quantum benchmark and execution data;
• private and enterprise-controlled AI deployments.
SynapseX
SynapseX is SoftQuantus’ AI and scientific intelligence platform. It is designed to support private AI agents, enterprise language models, scientific workflows and the integration of intelligent systems with internal applications, computational resources and research data.
The platform investigates how AI agents can plan, execute and verify complex workflows involving cloud systems, HPC resources, scientific applications and quantum computing infrastructure.
REPRODUCIBLE AND VERIFIABLE COMPUTING
A central research objective of SoftQuantus is the development of computational systems in which results can be independently analysed, reproduced and verified.
The group researches evidence-generation mechanisms capable of recording:
• source code and software versions;
• models and algorithm configurations;
• circuit compilation and transformation stages;
• backend and hardware information;
• execution parameters;
• measurement results;
• benchmark metrics;
• resource consumption;
• computational cost;
• provenance and integrity information.
This approach is intended to support scientific reproducibility, independent benchmarking, regulated environments and enterprise adoption of advanced computing technologies.
RESEARCH COLLABORATION
SoftQuantus is open to collaboration with universities, research laboratories, quantum hardware providers, HPC centres, cloud providers, deep-tech companies and independent researchers.
Potential collaborations include:
• joint scientific research;
• quantum and hybrid algorithm development;
• benchmark design and execution;
• access to emulators, simulators and quantum hardware;
• AI and LLM research;
• scientific software development;
• industrial proof-of-concept projects;
• research publications;
• European and international research programmes;
• validation of quantum and AI technologies in real operational environments.
Researchers, universities and technology partners interested in collaborating with SoftQuantus may contact:
