Meet the Founders: o-x.ai

Finalists in the TCF Market Research Innovation Awards

As part of our ‘Meet the Founders’ series, we introduce you to the visionaries behind some of the most innovative projects reshaping the industry.

Today, we’re excited to feature ox-ai, one of the Tony Cowling Foundation Market Research Industry Innovation Award 2024 finalists. We spoke with Director Zsolt Kiss, to learn more about the ground-breaking work his team is doing.

For our readers who might be unfamiliar with your work, can you provide a brief overview of what your project/service/platform does and the problem it aims to solve?

At o-x.ai, we specialise in applying AI and data science to research. Our flagship product, signAl™, is a Natural Language Processing engine that rapidly processes large volumes of qualitative data, extracting insights that would otherwise be too time-consuming or costly to obtain manually. We aim to streamline data integration, analysis and reporting, empowering researchers to focus on the high-value aspect of the process: insights.

What was the inspiration behind this project? Was there a particular moment or experience that sparked the idea?

The idea for signAl™ came from a challenge we encountered: we were faced with the daunting task of extracting information from over 44,000 pages of operational documents. Doing this manually using traditional qualitative methods was not only overwhelming but practically impossible within a reasonable timeframe and budget. We needed a way to speed up the process while maintaining robust quality standards. This meant ensuring the consistency, traceability and repeatability of the process; in other words, not building an AI black box. As such, we developed a tool that relies heavily on social science methodology but combines it with the power of cutting-edge AI.

Every startup has its ups and downs. Can you share one significant challenge you have overcome and a triumph that you are particularly proud of?

One of our biggest challenges has been overcoming scepticism around using AI for qualitative data processing and interpretation. Some of our clients and collaborators found it hard to believe that AI could deliver meaningful insights from complex, unstructured data, especially given the limitations other companies had encountered.

To address this, we undertook Proof of Concept projects, demonstrating the effectiveness of our "directed AI" approach, which involves applying specific AI tools to discrete tasks under the direction of our expert team. This meticulous process allows us to show, rather than just tell, what AI can achieve when used appropriately. By running PoCs across different projects and contexts, we provided transparent results, highlighting not just the capabilities but also the limitations of AI.

The turning point came when a particularly sceptical collaborator participated in a PoC with us. Through the process, they observed how AI, guided by human expertise, could extract consistent, actionable insights from large volumes of scanned documents. Seeing these results first hand changed their perspective entirely. Not only did they come to trust our approach, but they also became one of our strongest promoters, advocating for AI-driven qualitative analysis within their own network. This transformation reaffirmed the power of our method and the importance of transparency in building trust in AI.

Where do you see your project in the next five years? Are there any new features or innovations you're particularly excited about?

The future is already here, and it’s transforming the research industry. While I don't foresee existential threats, the way we work will undoubtedly change. Our product, signAl™, is just the first step. The long-term vision is to fully integrate across all data types – quantitative, qualitative, social media, client data – essentially allowing organisations to extract insights from their entire data universe.

At the moment, research often occurs in fragmented streams: quantitative data is processed separately from qualitative data, and social media data is isolated from both. Insights are then triangulated after the fact. But what if we brought triangulation to the beginning of the process? Why not integrate all these raw data sources upfront and extract insights from a complete view of the available knowledge?

signAl™ is the key to unlocking that potential. By addressing one of the largest hurdles – processing qualitative data at scale – it makes such integration possible. This shift will allow organisations to view their data holistically, generating richer, more actionable and more precise insights.

I envision signAl™ evolving into a tool that empowers businesses to move beyond traditional methods by merging all their data into one cohesive analytical process. This innovation won’t just enhance research — it will expand the boundaries of our industry into what today is the preserve of operational consulting. A holistic approach will position us to provide deeper strategic recommendations that go beyond data analysis, entering the domain of data-driven consultancy.

Beyond the business aspect, what personal motivations or passions drive you in this venture?

I'm deeply passionate about innovation and methodology. I believe in making the best use of the data we already have: it’s more efficient and less wasteful. There’s so much data collected, especially in social research and international development, but often it's underutilised. My aim is to change that by applying AI in ways that enhance efficiency and cost-effectiveness.

Looking back, is there something you wish you'd known when you started?

While it didn’t really come as a surprise, the effort required to balance the probabilistic nature of some of the AI algorithms we use with the precision dictated by our focus on social science methodology was definitely challenging. We knew AI could handle large-scale data processing, but ensuring that the outcomes were not only fast but also accurate, consistent, and replicable took more fine-tuning than expected. However, this process of exploration did teach us how to best combine quality and methodological rigour with speed and scale.

In an ever-evolving tech landscape, how do you stay ahead of the curve in terms of features and offerings?

We stay ahead by constantly looking forward and anticipating where technology is headed. With the rapid development of new tools, we prioritise flexibility, ensuring that we aren’t tied to any one tool or method. Our strategy is to stay nimble, adapt to the best available technologies, and remain open to innovation. This allows us to seamlessly integrate new advancements into our offerings, ensuring that we continue to deliver cutting-edge solutions that meet the evolving needs of our clients.

Every startup has a unique culture and set of values. Can you describe your company culture and how it contributes to your success?

Our culture thrives on a foundation of innovation, integrity and responsibility. We believe in getting things done not just quickly, but precisely – always on time, with honesty and transparency at the forefront. There’s a strong focus on collaboration, where we blend technological advancement with a human-centered approach. Every challenge is seen as an opportunity; we believe that no problem is unsolvable. This mindset fosters an environment where data scientists and social researchers work together to deliver insights that are impactful, actionable and rooted in both science and experience. It’s this commitment to innovation and quality, paired with our dedication to detail, that enables us to consistently meet the needs of our clients while staying true to our values.

Founding and running a startup is notoriously demanding. How do you and your team manage work-life balance, and how do you ensure your team does as well?

We prioritise flexibility and well-being above all. By leveraging AI in our operations, we reduce manual workloads, allowing our team to focus on high-value tasks. With team members spread across time zones, we offer the freedom to choose work hours that suit them, ensuring balance by tracking work hours and empowering everyone to say no when necessary. This keeps the team motivated, rested and engaged. Personally, running and exercising help me maintain balance, but this is definitely not a 9 to 5 job.

What would you say is the single most compelling reason people should choose your platform/service?

Quite simply, we deliver better insights, faster – and that’s no exaggeration. The service we provide excels at processing complex, unstructured data such as documents, administrative records, interviews, open-ended survey responses, and even audio, video, and social media. We do this at scale, identifying themes, generating summaries, narratives, or transforming qualitative data into structured datasets ready for further analysis. This combination of speed and depth allows organisations to act on insights quickly and confidently.