In this interview, AZoQuantum speaks with Dan Holme, Co-founder and CEO of Qoro Quantum, about advancing the quantum algorithm landscape through intelligent orchestration, scalable simulation, and hybrid computing infrastructure.
Could you please introduce yourself and Qoro?
I’m Dan Holme, co-founder and CEO of Qoro Quantum. The name “Qoro” comes from the Italian word for choir or chorus, which reflects what we do. We build an orchestration and simulation layer for quantum computing.
The industry is moving toward heterogeneous computing environments, combining CPUs, GPUs, simulators, and quantum processors. These systems are all essential for solving complex scientific problems, but stitching them together into a working workflow is incredibly difficult. It often requires months of custom integration and testing.
We founded Qoro to solve that problem. Our platform provides a unified software stack that is both hardware-aware and algorithm-aware. Instead of focusing purely on infrastructure, we look at the application itself and determine how best to distribute and execute it across available resources.
Image Credit: CineVI/Shutterstock.com
What is Solo?
Solo is our first cloud-based product and an accessible entry point into our platform.
Users can sign up, obtain an API key, and begin working immediately using our open-source SDK, Divi. Divi decomposes the problem, and Solo handles execution, orchestration, and infrastructure management behind the scenes.
At its core, Solo is designed as a simulation-first environment. We are not trying to be another gateway to quantum hardware. Instead, we act as an intelligent infrastructure orchestrator that simplifies execution and enables scalable experimentation without the complexity of managing underlying systems.
How would you explain the core value of Solo to a senior engineering leader skeptical of adding another cloud service?
I would not position Solo as just another cloud service, because it fundamentally does something existing cloud platforms cannot. Think of Qoro as a compression and parallelisation layer that sits on top of existing infrastructure. Teams can build once and run everywhere, while also leveraging all available compute resources simultaneously.
This removes the need to rewrite algorithms for different backends and allows late-stage flexibility without re-engineering the entire workflow. It also unlocks significant gains in efficiency. For example, a QAOA workload that would cost over $50,000 when run naively was reduced to around $360 using our approach.
Beyond cost, we are also addressing developer productivity. Automation and orchestration reduce the burden on highly specialized talent, which is in short supply in quantum computing. Ultimately, the goal is a world where users of quantum computers do not need a PhD in quantum computing to build their own algorithms.
Video Credit: Qoro Quantum
What does Qoro’s multilayered parallelisation mean technically, and how does it improve scalability?
Multilayered parallelisation refers to combining multiple independent parallelisation strategies across different layers of the stack.
At the application level, problems can often be decomposed mathematically into smaller components. These components can then be executed in parallel. At the algorithm level, elements such as observables can also be parallelised. At the hardware level, we can pack multiple logical circuits onto a single QPU, improving throughput. On top of that, simulation adds another layer, where parts of the workload can be offloaded to classical systems that are more cost-effective.
The key is that these layers compound rather than compete. Together, they significantly improve scalability and efficiency, enabling workloads that would otherwise be impractical to execute.
How do Divi, Composer, and Maestro divide responsibilities within the platform?
Qoro Quantum's platform separates quantum software into three layers: problem definition, orchestration, and simulation.
Divi is an open-source Python library where developers define quantum programmes. It serves as the entry point to the platform, handling problem construction and parallelisation logic.
Composer is the distributed substrate layer. It takes programmes from Divi and prepares them for execution: circuit packing, observable grouping, device selection, and job scheduling. Composer inspects backend topology and qubit availability, then allocates workloads across available resources. In published benchmarks, it achieved an 83% cost reduction on IBM's Heron 2 processor and a 96% reduction in circuit count through automated observable grouping.
At the execution layer, workloads route to Maestro for unified simulation and/or to third-party compute (QPUs, HPC, cloud). Maestro integrates statevector, MPS, Clifford, tensor network, and p-block backends under a single interface, automatically selecting the most efficient simulator across CPU and GPU. Its runtime predictions feed back into Composer's scheduling decisions, closing the loop between simulation and hardware execution.
Developers write programmes once in Divi. The platform handles everything downstream. We structured the platform this way to separate concerns while maintaining flexibility. Each layer has a clear role, but together they form a cohesive system that abstracts away complexity for the user.
Keep this interview for later - download the PDF here
What did your benchmark running a 250-node QAOA max-cut program reveal?
We ran a benchmark involving a 250-node QAOA max-cut problem with around 9,000 jobs executed in under 10 minutes. This highlighted several key insights.
First, classical simulation bottlenecks become very apparent at scale, particularly when workloads are not efficiently distributed.
Second, naive execution approaches simply do not scale. Without orchestration and parallelisation, the cost and time quickly become prohibitive.
Where Solo delivers the biggest gains is in how it distributes and compresses workloads. By combining multiple optimisation techniques, we can dramatically reduce execution time while maintaining performance, enabling experiments that would otherwise be infeasible.
How does Solo help move research-stage quantum algorithms closer to deployable solutions?
One of the biggest challenges in quantum computing is closing the gap between research and real-world deployment. Solo addresses this by providing a consistent execution environment where algorithms can be tested, scaled, and optimised without needing to be rebuilt for different systems.
By abstracting the underlying infrastructure and supporting hybrid execution, teams can spend more time refining the algorithm itself instead of managing technical complexity. That helps speed up the move from proof of concept to solutions that can integrate with enterprise IT systems. In practical terms, it shortens the path from experimentation to real application.
What were the hardest challenges in unifying CPUs, GPUs, simulators, and QPUs?
The biggest challenge is treating fundamentally different systems as a single logical machine while managing latency, contention, and cost. Each system has different performance characteristics, communication overheads, and constraints. Coordinating them in real time requires deep awareness of both the hardware and the algorithm.
Another challenge is ensuring efficient resource allocation without introducing bottlenecks. If one part of the system lags, it can impact the entire workflow. Solving this required building a platform that can dynamically adapt execution strategies based on both the workload and the available infrastructure.
Where do you see the industry going?
The industry is clearly moving toward hybrid and distributed computing models. Quantum computers will not operate in isolation. Instead, they will be part of larger systems that include classical computing and simulation. The key challenge will be making these systems usable and accessible.
We believe orchestration and automation will play a central role in enabling that future, allowing more users to engage with quantum computing without deep specialist knowledge.
Download the PDF of the interview here
About the Speaker
Dan Holme is the CEO of Qoro Quantum, where he leads the mission to build the unified software stack for distributed and hybrid computing. Coming from a strong IT background - including previously leading global industry and consulting partnerships at Cisco, where he also served as a quantum technologies specialist - Dan brings a pragmatic, enterprise-first approach to the quantum industry. He focuses on helping HPC and enterprise organisations orchestrate heterogeneous quantum and classical resources to maximise return on their quantum investments
Disclaimer: The views expressed here are those of the interviewee and do not necessarily represent the views of AZoM.com Limited (T/A) AZoNetwork, the owner and operator of this website. This disclaimer forms part of the Terms and Conditions of use of this website.