Marta Garcia is CFO of Multiverse Computing, a Spain-based AI company focused on making artificial intelligence more efficient. The company uses technology derived from quantum computing to compress AI models, helping customers reduce costs, lower energy consumption, and deploy AI on a wider range of devices.
Since joining Multiverse Computing in 2019, Garcia has helped guide the company through a period of rapid growth. Today, its customers include organizations such as Iberdrola, Moody’s Analytics, the Bank of Canada, and Telefonica. In 2025, the company was named one of CB Insights’ 100 Most Promising Companies in AI.

In this conversation, Garcia shares her path from investment banking to the CFO role, discusses the challenges facing AI companies, and explains how Multiverse Computing is helping customers run AI more efficiently.
Can you tell me a little bit about your journey? How did you become a CFO?
My background is in corporate and investment banking. I’m Spanish, but I moved to London when I was 22 and worked there for almost 10 years before moving back to Spain.
At one point, I moved back to the Basque Country, where my family is from, and saw a position advertised at Multiverse Computing. I met with two of the founders, and it was more of a matchmaking process than anything else. I did not know much about quantum computing or AI at the time, but I knew I wanted to work with them and help make the company successful.
When I joined, we had seven or eight employees. Today, we have around 500.
Do you think there’s any traits that are important for a CFO?
If you cover both finance and fundraising, the role can sometimes feel closer to a sales position than a traditional finance role.
You need to understand the product, the different departments within the organization, the culture, the vision, and where the company is heading. Then you have to communicate that whole picture to investors.
Networking skills, the ability to understand other people’s perspectives, and strong interpersonal skills are all very important.
What would you say success would look like for Multiverse Computing over the next 12 months?
We are currently closing our latest funding round, so success is linked to what we want to achieve with that capital.
There are two key markets we are targeting. One is gigafactories, particularly in Europe, where there is significant investment and interest. The other is edge deployment, where our technology allows highly compressed AI models to run locally on devices.
If we can establish ourselves in those markets and move quickly, the opportunity is enormous.
Which business outcomes matter most to customers when you’re providing AI model compression?
The biggest areas are cost and deployment.
AI is expensive, and the more use cases companies implement, the more those costs increase. Our models are more efficient and consume less energy, which helps reduce those costs.
We can also compress models enough to deploy them directly onto devices, allowing them to operate locally rather than relying entirely on the cloud. That could be a laptop, phone, electric vehicle, satellite, drone, or other device.
What are enterprise buyers really prioritizing in AI? Is it performance, efficiency, data sovereignty or something else completely different?
It depends on the customer.
For governments, pharmaceutical companies, and financial institutions, reducing infrastructure costs can be a major priority because they need fewer GPUs to run the same models.
For device manufacturers, the focus is often on deploying highly efficient models that maintain accuracy without consuming excessive memory or computing resources.
In Europe, data sovereignty is also an important consideration, particularly around AI infrastructure and gigafactories.
What do you think the biggest strategic challenge is facing AI companies?
Cost remains one of the biggest challenges.
As organizations expand their use of AI, costs continue to increase. At the same time, many companies are still trying to determine where to begin and which use cases will deliver the most value.
Many organizations want to adopt AI but are still working out how to focus their efforts and where to start.
What would you say is the biggest challenge that you face as a CFO?
AI moves incredibly fast.
There are constantly new models, new technologies, and new developments entering the market. As a CFO, you need to stay informed, understand what is happening, and be able to explain those developments to both existing and prospective investors.
The metrics used to evaluate AI companies are also evolving, so there is a constant need to understand how the market is valuing businesses and how those benchmarks are changing.
Is there anything exciting happening in your role right now?
The funding round is the most exciting thing happening right now because it occupies most of my time.
Beyond that, there are exciting developments taking place across the company. Our research team is exploring how very small language models might eventually run on quantum computers. It does not work particularly well today, but if it does in the future, it could become the first language model running on a quantum computer.
There are also new developments happening almost every day, which makes the environment incredibly exciting.
Has anyone ever given you any advice that really stuck with you and helped you in your role as a CFO?
One piece of advice that has always stayed with me is not to be afraid of making mistakes.
That’s how you learn. I would rather apologize for trying something than avoid doing something that I genuinely believe could be the right decision.
Is there anything fun that you do to prevent burnout?
Running used to help me a lot, and I would like to spend more time exercising again.
At the moment, though, most of my time outside work is spent with my family and my six-year-old daughter. Family is what keeps me busy and entertained outside of work.
What would you say is your proudest moment as a CFO?
Closing our Series B round was a proud moment.
I remember bringing together a group of investors who all believed in the same vision. There was a real sense of energy and excitement in the room, and seeing people align around what we were trying to build was very rewarding.
Is there anything that’s inspiring you today?
My daughter is a huge source of inspiration.
I am also passionate about raising awareness around menopause and perimenopause. I think there has historically been a lack of discussion around these topics, and many women do not always receive the support or information they need.
Making those conversations more visible and helping people better understand the challenges women face during that stage of life is something I care deeply about.
Are there any new projects you are currently working on?
One recent development is a new product we have built in-house.
It is a small language model that fits on a device and acts as a router, deciding whether a workload is simple enough to be handled locally or whether it should be routed to the cloud. The system connects local AI with cloud-based AI and works with models hosted through our AWS integration.
The product links together AI running on devices, on-premises environments, cloud infrastructure, and gigafactories, creating a unique bridge between those different deployment models.
Marta Garcia is CFO of Multiverse Computing, a Spain-based AI company focused on making artificial intelligence more efficient. The company uses technology derived from quantum computing to compress AI models, helping customers reduce costs, lower energy consumption, and deploy AI on a wider range of devices.
Since joining Multiverse Computing in 2019, Garcia has helped guide the company through a period of rapid growth. Today, its customers include organizations such as Iberdrola, Moody's Analytics, the Bank of Canada, and Telefonica. In 2025, the company was named one of CB Insights' 100 Most Promising Companies in AI.
In this conversation, Garcia shares her path from investment banking to the CFO role, discusses the challenges facing AI companies, and explains how Multiverse Computing is helping customers run AI more efficiently.