© ROOT-NATION.com - Use of content is permitted with a backlink.
If you’re a regular visitor to our website or follow us on social media, you may have already heard that we recently introduced a virtual author – a bot named Ruta AI – to generate news content automatically.
Today, however, we decided to take a closer look at the development of cutting-edge AI solutions. We’re sharing a short interview with the CEO of Unpress AI, a Ukrainian startup that developed the tool recently implemented on our website.
Interview with Alex Shpantsev, CEO & founder of Unpress AI
What is your professional background? Where did you start and how did it influence the creation of Unpress AI?
For over eight years, I worked in sales and marketing for blockchain projects. I started during the ICO boom as a marketing assistant at a startup, then moved on to a full marketing role. Eventually, I founded my own agency focused on blockchain product marketing. I thought I had a solid understanding of how startups operate – I had worked in them, with them, and helped launch dozens of projects. But when the time came to build something truly my own, I realized it was a completely different game. My previous experience was helpful, but I had to relearn everything: making final decisions, leading a team, and taking full responsibility for every step. It was no longer about offering advice from the sidelines – it became a personal all-in commitment.
The idea to create Unpress AI came after my discharge from the Armed Forces of Ukraine.
It was a particularly meaningful period – one where you begin to reflect on the value of time and the impact of your decisions. I came across a study on the use of AI in newsrooms, and it left a strong impression. It pointed out that despite many attempts to integrate artificial intelligence into media workflows, there were very few successful cases. The rate of effective implementation was low, even though the potential was obvious.
That got me thinking – why? I began digging deeper and quickly realized the problem wasn’t a lack of technology, but a lack of structure. The ideas were there, but what was missing was a cohesive system – something that could connect all the key stages: from news monitoring and fact-checking to content creation and publication.
In February 2023, I began shaping the design of the solution that later became Unpress AI. By the summer of 2024, we had already moved into development, and in December, we reached our first pilot clients. This wasn’t just about technical development – it was about a deep conviction that we could change the approach to news.
Why did artificial intelligence become the basis of your decision, even though you had no experience with it before?
Throughout history, humanity has experienced multiple technological revolutions – the wheel, electricity, the assembly line, and automated manufacturing. None of these innovations “took away” jobs. Instead, they eliminated routine tasks and physical strain, opening up new opportunities for creativity, analysis, and management.
Artificial intelligence is the next logical step. I believe it won’t replace humans, but rather shift the balance – allowing us to focus on what’s complex, interesting, and strategic. That’s why the adoption of AI solutions in any field is inevitable. In this new world, there are two paths: either create these solutions or stand aside and wait for someone else to do it. I consciously chose the former.
Did you have any doubts or fears before launching Unpress AI and how did you overcome them?
Of course. For me, it’s like a wave on an oscilloscope screen – the level of internal belief in the project constantly fluctuates. And I think that’s completely normal. It’s a natural state for most people creating something new. Even if at some point that belief starts to dip – I keep moving forward, because I know: it’s part of the process. The belief doesn’t disappear – it just changes amplitude, but the vector remains the same.
To achieve results, it’s not enough to knock on dozens or even hundreds of doors. You have to do it with the same enthusiasm – whether it’s the first time or the hundredth. I believe that’s what truly defines a person’s readiness to see the journey through.
Like most startups, we face the challenge of funding. In the past, I deliberately avoided seeking investments – I wanted to maintain full control and implement my vision without compromises. But over time, I realized: if you want to move quickly and on a large scale, it’s a basic necessity. The speed of implementing ideas is critically important. And now, I’m actively looking for partners who share the values of Unpress AI and are ready to support the next leap in our development.
Who is your main target audience? For whom are you creating a product?
Our main audience is news editorial offices in all their forms. This is a B2B model, focused on professionals who create or curate news content on a daily basis. We offer a comprehensive solution that can replace a patchwork of separate tools – dozens of individual systems that editorial teams have to manually combine, from monitoring and translation to separate fact-checking systems.
Essentially, anyone who publishes news in any format can benefit from collaborating with us. At this stage, we are capable of removing up to 80–90% of the routine workload from editorial teams – tasks related to information search, verification, translation, and basic content creation.
In the long term, we also see great potential in the B2C direction. However, at this stage, it is critically important for us to maintain tunnel vision – a focused approach on the core product. We don’t want to spread ourselves too thin because we are creating a solution that has the potential to transform the very infrastructure of the news industry. And for that, we need to concentrate on every detail.
Do you already have real customers or cases of using Unpress AI? What do they say?
To date, we have nine pilot clients – and this is more than enough for deep testing and comprehensive product refinement. We deliberately limited the number of pilots to work with feedback as thoroughly as possible and to draw clear conclusions: what the market really needs, what the solution should look like, and how it should operate.
The feedback has been overwhelmingly positive. Of course, there were nuances that we, having no previous experience working in editorial teams, might not have considered. But overall, there hasn’t been a single dissatisfied client.
Interestingly, during the testing process, we gained a few unexpected insights. For example, local media is a completely different reality. There are far fewer traditional sources, fact-checking is more difficult, and social media channels of local authorities or influencers often serve as the main informational reference points – sometimes these are public figures or activists who aren’t part of the mainstream media landscape. By the way, we are actively preparing to launch social media monitoring based on this very request.
What is the main value of your product for the customer? What do they actually get?
In a sense, we operate like a news conveyor belt. Somewhere in the world, a news story emerges – the system instantly finds it, verifies its relevance, and breaks it down into facts. Each fact undergoes an internal verification process using our algorithms. Based on these verified facts, the artificial intelligence generates a unique piece of content that aligns with the specific settings of the client – such as style, length, tone of voice, structure, and more.
The content is then automatically sent directly to the client’s CMS. They only need to make minor adjustments – and it’s ready for publication. As a result, we significantly speed up the editorial process, eliminate routine tasks, and, accordingly, drastically reduce both time and financial costs.
In business, it often comes down to the formula “time is money,” and we help save both resources at the same time.
How do you work with sources and check their reliability?
We have an internal trust scoring system – a dynamic trust rating for sources. This is not a static evaluation, but a system that is constantly updated based on comparison results.
When our system finds multiple publications on the same topic with conflicting facts, the trust rating helps determine which source is more reliable. If a source repeatedly spreads fake news or dubious information, its rating is automatically lowered. This affects the priority given to the source when creating a news piece for the client.
Thus, the more a source “proves” its accuracy, the more the AI trusts it. And, as a result, the final output for the client will be more accurate and reliable.
How scalable is the system? What languages, regions or topics does it already cover?
Currently, we are testing the monitoring of around a thousand sources – this number constantly changes as the system evolves. By the end of the year, we plan to expand to over 5,000 sources, and in the longer term, potentially many more. Additionally, we are preparing a separate module for monitoring social media. So, the total number of information gathering points could easily exceed 100,000.
As for languages, AI sees no barrier. Language is not “special” in the traditional sense; it is simply another form of neural connections. Therefore, the system works equally effectively with any language – searching, reading, analyzing, and creating.
And most importantly – scaling in this field is technically very simple. Compared to many other niches, expanding to new markets, topics, or languages is a matter of data structure and model adaptation, not building everything from scratch.
What is your vision for the future of Unpress AI? Where are you going?
I sincerely dream that in 3-5 years, newsrooms will no longer have to manually perform all the monotonous and painstaking work. We want to provide them with a tool that will allow them to focus on what matters – original journalism, investigations, and analytics. That is, on what machines will never be able to fully create.
On our roadmap are new modules, including full integration with social media, which today is the main source of news trends. Our goal is to automate this layer as well, making data processing from social networks just as systematic and reliable as it is with traditional sources.
For me, Unpress AI is a continuous cycle of victories, mistakes, and improvement. Artificial intelligence is simply the next step in technological development. But the most important thing is not the tool itself, but how we use it to change the approach to creating truthful, fast, and high-quality news.
