Yuval Boger interviews Vishal Chatrath, CEO and co-founder of QuantrolOx, a quantum control software company focused on automating qubit tuning and calibration. They discuss how automation accelerates chip characterization, supports scalable manufacturing, and feeds into real-time calibration and error correction. The conversation covers competition in quantum control, open architectures, fundraising challenges, and what it takes to industrialize quantum hardware.
Transcript
Yuval: Hello Vishal, thank you for joining me today.
Vishal: Very nice to be here, Yuval. The last time we spoke was in early 2023, so I’m absolutely delighted to be back after nearly three years.
Yuval: It feels like centuries ago, but remind the audience, who are you and what do you do?
Vishal: My name is Vishal Chatrath. I’m the CEO and co-founder of a company called QuantrolOx that spun out from the University of Oxford. We provide software to automate the tuning of qubits.
Yuval: The field has gotten probably a little bit more crowded since we last spoke, right? I am aware of several other companies that claim to have similar things or control software or control software and hardware. Who do you see as the main alternatives to your company and how do you differentiate?
Vishal: I think that’s a very good question. I think one of the key differentiators that we’ve had since our inception is that we were unique among people building control software or control software for automation from the fact that we were experimentalists who were building software for experimentalists. Right.
There are some other companies in the field, all fantastic companies, but they do come from a theory background and they have a very different take on what it means to do quantum control. And also because the field is so early, we don’t have a language with high enough gate fidelity to be able to talk about different kinds of control challenges.
An example of which I will give is that we have one of our customers in Taiwan where our assumed competitors — I Ithe companies here — but they are in that same customer and the three companies are doing three completely different things. But I think that really explains that there is a lot more nuance here than is understood.
So our specific focus is really about whenever you take a new chip out of the oven, how do your basic tests, is it any good for an operational environment? To be able to do that, you get a fresh chip out, you have to first of all even check if the chip is any good. So that means you have to bring it up, check all the resonators, all the qubits, everything’s working.
Go through all the single-qubit experiments, Rabi, Ramsey, et cetera. Then go through all the two-qubit experiments to check if those two things entangle.
Without our software, the way it is done by everybody else, that entire process can take two or three PhDs up to a week to go through all these experiments because there are about 20 experiments to be able to do.
What we showed in March of 2025, live at the APS March Meeting, was that you can press a button and go and have a cup of coffee or if you eat really fast you can have your lunch and you can come back in under 25 minutes, this is done.
So what was a PhD-level skill, we have made it into a mass-market phenomenon that allows industrialization of quantum chips.
For people like us who are maybe a bit more older in the industry, because I did my degree in the early 90s in electronics or in microelectronics, we refer to these as EDA tools, electronic design automation, right. So there is the design element and there’s automatic testing of semiconductor chips.
So we are the quantum equivalent of an EDA tool for quantum and it’s the fundamental thing that you need to be able to move towards industrialization of quantum chips and quantum components. That is where we are unique.
Yuval: I work for a company where we don’t take things out of the oven. I work at a neutral atom company. And so some of the terms that you mentioned are a little bit foreign to me. So help me understand, if I create a chip and I take it out of the oven and I test it with your tools and it’s fine, how often do I need to do that again? I mean, do I need to do that every time I manufacture a new chip? And if so, right now people don’t make hundreds of thousands of quantum computers, so how often would I need to use your software?
Vishal: Very good question. So we are in that entire basically R&D phase, right?
So, you know, just to give you a perspective, let’s go to the world of semiconductors, right? So the first transistor was made in 1949 and it was the size of my thumb, right? You can basically go to Google and you’ll see kind of replica images. So I’m not joking. It was literally the size of my thumb.
And by 1961, from 1949 to 1961, we had the first integrated circuit. And that had 12 transistors. In 1971, the first microprocessor had 2,300 transistors, versus today the Cerebras chip with three or four trillion transistors, right?
So the EDA tools market in 1949 started developing year on year as the complexity of chips has grown and the volume of chips has grown. In 2025, the EDA market was 200 billion.
Coming back to here, how often do you need to use it? We will need to go through tens of device generations before we find scalable chips, as in the case of semiconductors.
In every generation, you have to go through hundreds of devices, and in each hundreds of devices, you have to do thousands of experiments to get enough data to move to the next one.
So for example, our customers in Taiwan, they were doing five to ten chip characterizations a year before they started to use our software in April 2024. As of now, they’re characterizing a new chip every day. And that’s not enough.
They are updating their fabs so that sometime in 2026, they’ll be characterizing 40 samples a week. So that’s 2,000 samples a year.
So in any development, especially when it comes to materials, there’s no shortcut. You have to do this by brute force and you have to really do thousands upon thousands of samples.
But having said that, you do it for the R&D QA process and then feed back into your R&D loop to improve the materials.
But it’s essentially the same qubit automation that you also need to use during the runtime of your quantum computer because your qubits are very special.
No two qubits are the same, and the behavior of the qubits drifts over time. And I’m not talking about drifting over months, it literally drifts over hours.
So you have to constantly at millisecond or nanosecond intervals ping the qubits to understand where the characteristics are or where the energy level is from what is ideal.
Based on that, you literally have this thermostat — a fancy thermostat I would like to call it — to fire back microwave pulses to energize the qubit to the right state.
So this is a constant process. So you never really stop characterizing, because if you stop characterizing, you don’t know how to basically energize the qubit.
Yuval: But this goes back into my competitive question, because I think that some of the other companies at least position themselves as doing these real-time calibrations while the systems are interlaced with the actual calculations of the system. And you started by saying we’re focusing on the manufacturing process and the characterization, but then do you also do the real-time calibration?
Vishal: We also do that. So for sure there’ll be an overlap, right?
But what makes us — I don’t want to use the word better — but yes, is that we have been involved with the QPU or with the qubits since their birth.
So ultimately, as part of your workflow, ideally you want to have the same tools for which you have trained your algorithms to handle them at the manufacturing R&D and QA process, and you want to keep using the same training and the same algorithms because that will just give you more operational efficiency.
And again, I take my experience back to my early days as an engineer. After I left the National University of Singapore, where I was doing device fabrication and material characterization, I worked in fabs and then moved to industry.
When I was working for HP, I used to build test systems which were first used by R&D, and then the same software — toned down and with the UI cleaned up — was used in manufacturing.
Because these are very complex products. You can’t say, okay, I assume that if you have to make it work at runtime you’ll get a chip for which you have no background.
We already know how the chip behaves because we are in the manufacturing environment, so why reinvent the wheel?
Yuval: Does the product need to change now that people care more and more about logical qubits and quantum error correction?
Vishal: I think for sure it has evolved.
When we introduced the product in 2024, we had a perspective that you just press a button, the whole automation happens and that’s it.
But then we realized that we were selling to an audience trained to be skeptical — engineers, physicists — and people said, okay, you’re doing 20 steps super fast, can you break these up? Can we edit the workflows and adapt them?
So we introduced the SDK.
In parallel, we were approached by people working in quantum error correction, and they said for QEC we need the most up-to-date state of the qubits. Keeping qubits tuned reduces overhead.
We are working on APIs so that QEC software can tap into us.
After the SDK launch, people can now use QUA, Quantify, Q-Codes, on Keysight, Zurich Instruments, Qblox, or even homebrew electronics, on superconducting QPUs.
We are also working on spin qubits — I can’t name the company yet — and later other microwave-based and laser-based modalities.
Yuval: When we spoke three years ago, your product was 100% software. Is that still the case?
Vishal: Until a few months ago, yes.
But in emerging markets, customers didn’t have systems. They visited our lab, saw our open-architecture quantum system — not a computer — and asked us to build it for them.
So we packaged the components we already use. Think of it as a Raspberry Pi for quantum research. We don’t do compute. We stop at the physics layer.
Yuval: Running a quantum company can be a roller coaster. When you think about the past three years, what were the high points and low points?
Vishal: High points happen almost every day.
Low point was 2025. We went against closed-architecture orthodoxy with open architecture. Development took longer. Fundraising was brutal.
In the first nine months of 2025, we signed two customers. In Q4, we signed six. This year, we’ve already hit seven figures in sales.
2025 was brutal.
Yuval: Which of your original assumptions proved to be wrong?
Vishal: I assumed nobody would want to manually run 20 experiments. Customers wanted control. We had to re-architect the software. It took about 12 months. Very humbling.
Yuval: Do you see your products being more popular in certain parts of the world?
Vishal: From a solid-state perspective, the best fabs are Japan, Taiwan, South Korea, and the US. That’s why our first customer was in Taiwan.
Now we have customers in Europe, the US, academia, and industry. We deliberately diversified geographically and by segment.
Yuval: As our time comes to a close, three years ago you wanted dinner with Richard Feynman. Who would you want to have dinner with today?
Vishal: I think I was thinking about that, the whole question the last time we spoke, and then I thought this whole question would kind of come up again, and I was thinking that among all the crazy people that we’ve had in the world of quantum physics.
And maybe I would stress that definition a bit and talk about even the early days of X-rays and all that period, right, to the early part of the 1900s.
I don’t think you have more bold and passionate person than Marie Curie and, you know, kind of Pierre Curie, who just for their passion of science, fear had no place.
Right, so the kind of risk that you take.
And I think they are probably the closest to defining the mindset of a deep-tech entrepreneur, right? Because all of us by our nature are just going into the unknown like nobody’s business with actually very little concern for, yeah, you know, what’s going to happen tomorrow.
And I think these are — so I think — and of course it’s not even remotely as dangerous as what they were doing, but to me they are the ultimate risk takers if there were any to advance the cause of modern physics, and without them we wouldn’t have any quantum physics.
Yuval: Vishal, thank you so much for joining me today.
Vishal: All right, thank you very much, Yuval.
Yuval Boger is the Chief Commercial Officer of QuEra Computing.