AI in action: Transforming operations, products, and careers

Though the Innovantage podcast is focused on technology and innovation in general, artificial intelligence remains the most widely discussed topic in the episodes. AI is everywhere, and it is impossible not to talk about it while discussing the changes that are happening around us.

This episode hasn’t become an exception. The podcast host and Sigli’s CBDO, Max Golikov, together with his guest, Martynas Kairys, discussed the real impact of AI in payments and business operations, as well as the possibility of turning experiments with AI into strong competitive advantages.

Today, Martynas is the General Manager at ZEDGE, an AI-driven growth leader, and an international keynote speaker.

But as he admitted, he had never expected to work in technology. When he graduated in 2000, computer science was already booming. However, at that time, he was convinced that it wasn’t for him. His academic path led him to the Danish Royal Naval Academy, where he became a naval officer, and later to a degree in economics in Lithuania. Such fields are quite far from tech.

Nevertheless, everything changed in 2012, when Martynas decided to launch a startup. The idea was to create an app that would help people build good habits. As Lithuania’s startup scene was just beginning to grow, he found himself at the center of this emerging movement. He began learning how to code in order to understand the needs of his team better. Though he never became a professional developer, this knowledge helped him bridge the gap between leadership and technical teams.

After closing the startup, Martynas moved into product and program management roles at Shift4. There, he oversaw developer marketplaces, third-party integrations, and sales systems. 

Over time, his focus shifted toward automation. Long before generative AI became mainstream, he was already exploring intelligent systems and studying AI at a strategic level.

When ChatGPT took off in late 2022, Shift4 created a dedicated AI leadership role, and Martynas stepped in to guide the company’s adoption of this technology. 

Driving innovation at Shift4

Shift4 is a global fintech company specializing in payments. It has its headquarters in the United States and its largest office outside the US is based in Vilnius, Lithuania. When Martynas joined the team, the office in Lithuania had just 40 employees. Today, it has grown to more than 800. The Lithuanian team now contributes across multiple areas, including product development, R&D, and customer support. The company can boast a diverse client base that ranges from restaurants and stadiums to casinos, hotels, and even such giants as Starlink.

Within Shift4, Martynas always implemented a proactive approach. Instead of just bringing ideas, he also built prototypes and tested them. He was often the person who started projects, like the developer marketplace and lead management system. They were built from scratch and further scaled into full teams and products.

Martynas also supported innovation inside the company. He launched Shift4’s first hackathons in Lithuania and later on a global scale. 

How AI can change services and internal operations

At Shift4, artificial intelligence plays a growing role in both customer-facing services and internal operations. The company develops solutions that help clients use its payment services more effectively and also streamline workflows for employees.

One of the most widely adopted tools is an AI assistant. It is a chatbot that answers practical questions for merchants, such as how to process a refund, add new employees, or troubleshoot common issues. Traditionally, users had to read lengthy guides or wait in line for support to get answers. Today, the AI assistant can deliver clear, step-by-step instructions instantly, which helps save valuable time.

The technology goes beyond simple troubleshooting. Clients can now ask quite complex questions, such as: “What were our latest sales of specific items compared to the same period last year, percentage-wise?” Now people don’t need to find the required data in spreadsheets and then perform calculations manually. AI can provide the analysis on demand.

In the internal operations, AI helps people with everyday questions that cover various aspects, from how to request vacation time to why a particular product feature was built. In a company with thousands of employees, automating these queries reduces constant interruptions for teams like HR or engineering and frees them for more strategic work.

Developers also benefit from AI-driven tools like GitHub Copilot. Such solutions help accelerate coding by suggesting and refining lines of code.

Beyond off-the-shelf tools from companies like OpenAI, Anthropic, Microsoft, and Google, Shift4 encourages employees to create tailored AI solutions. Marketing teams, for instance, can build custom GPTs that generate posts in the company’s brand voice, while developers use AI companions to explain code or review logic.

The culture of using AI

Speaking about his attitude to artificial intelligence, Martynas explained that, on the one hand, AI brings enormous potential. On the other hand, its application is related to risks that people will use it only as a copilot that delivers results just good enough to pass.

But over time, businesses and consumers will demand higher quality and drive expectations for better products and services.

For him, the impact of AI depends not only on what the technology can do, but also on who uses it. His grandfather, for example, adopted ChatGPT’s voice mode in Lithuanian to get advice on gardening. Despite being hesitant to pay for other services, he agreed to subscribe to ChatGPT because it solved a real need in a simple, accessible way.

In professional contexts, AI can be used in surprising ways. Some developers use it to generate code. But others value it more for having meaningful discussions about their work. It helps them refine their solutions. Others leverage it to communicate better. One developer shared that AI helps him present ideas clearly to colleagues and management. As an introvert, he can now better express his thoughts, which results in faster alignment and decision-making within his team.

The art of prompting

Martynas emphasized that the quality of AI output depends heavily on how questions are framed. A simple request like “Tell me a joke” produces something bland. But when you add just a few words and your request sounds like “Tell me a joke in Jimmy Carr’s style”, you will get an absolutely new result.

The same principle applies to professional use cases. If you ask AI for generic feedback on a sales letter, you will get neutral comments. When you prompt it to respond as a VP of Sales at a specific type of company, you will receive far more relevant insights.

Martynas advised users to refine prompts by adding context and even letting AI guide the process. One of his favorite tricks is to finish a request with: “What information do you need from me to answer better?” This forces the system to ask clarifying questions. Thanks to this, it can provide more accurate responses.

To test the power of prompting, Martynas organized a public experiment in Lithuania on ChatGPT’s two-year anniversary. Together with a communications expert, he challenged 12 professional writers to submit essays on the theme “Why I write and what motivates me.” Martynas then created his own 13 texts using ChatGPT. To perform this task, he used different prompts, sometimes seven pages long. They included detailed stylistic instructions and writing samples.

The experiment produced 25 texts in total, which more than 1,400 participants tried to classify as human- or AI-written. The results were surprising. People could barely tell the difference. One of the pieces was mistakenly classified as human-written by 76% of respondents. But in reality, it was generated by AI using the prompts prepared by Martynas.

The importance of AI in MVP development

Martynas believes one of the biggest misconceptions in technology is that companies need huge budgets to test new ideas. In traditional corporate settings, innovation can move slowly. With AI tools, however, teams can now build strong MVPs quickly and affordably.

Instead of spending months and significant resources, it’s now possible to launch a simple web app, create landing pages with polished copy, or even deploy clickable beta versions in less than a day. All this is available when you use relevant platforms. Martynas recalled a recent hackathon where participants produced a working prototype within just four hours. The same process would have taken an entire weekend only two years earlier.

This shift plays a huge role in the business space. While working on his app for developing good habits many years ago, Martynas pitched an early version using just a Google Sheets demo. It was crude, but it secured investment. Today, founders and teams can achieve significantly more and significantly faster.

Still, many companies are convinced that MVPs require heavy investment or outside consultants. External experts can help. But it is also worth paying attention to internal talent. Every organization has domain experts who can become real AI evangelists. Their deep industry knowledge and interest in AI experimentation can significantly accelerate innovation and help deliver solutions that will be better tailored to the business needs.

AI and regulation

Regulation in the AI world is one of the hottest questions today.

According to Martynas, regulation in the European Union is far stricter than in the United States and other regions. Oversight is essential in many areas, but excessive rules can slow innovation. But what is even more important, they don’t necessarily make systems safer.

The easiest way to ensure no breaches is to forbid everything, but that comes at the cost of progress.

Martynas emphasized that AI can already generate convincing text or code, but it often requires domain experts to assess whether outputs are reliable. For now, businesses in regulated industries are cautious, and AI is rarely implemented in core operations.

Moreover, we shouldn’t ignore the psychological barrier. People may be comfortable with autonomous cars today. But pilotless planes or AI-only doctors can still cause distrust. Despite improvements, AI still hallucinates, and when facing critical decisions, like surgery, humans want human oversight.

How to build a startup

With his solid business expertise, Martynas believes everyone should experience both building a startup and working in a corporate environment. Each path teaches different lessons. Startups demand agility, resilience, and creativity, while corporations provide structure and exposure to regulation.

However, timing is crucial. Launching a startup while raising a newborn, for example, can feel like running two startups at once. Family support and financial stability are key, as entrepreneurship carries risks and uncertainty. 

Contrary to the myth of the young tech prodigy, research shows that founders over 40 often build the most successful startups. Many of them transition from corporate roles and have both maturity and experience. 

With the rise of AI, there are even more opportunities for those who are ready to build their own projects. Solo founders can now achieve what once required entire teams. Even a single person with the right tools can build something extraordinary.

Ideas for a startup

Martynas advised future founders to create something they would personally use. Passion is crucial because most ideas already exist in some form, and without genuine interest, it’s easy to get discouraged.

Market fit is the number one reason startups fail. Many founders build “vitamins” (nice-to-have products) instead of “painkillers” that solve real, urgent problems. After market fit, the next challenges are financing and team building. Nevertheless, without clear demand, even great ideas can’t scale.

Professional development: How AI will change it

AI has shifted the value equation in the workplace. A single senior developer equipped with AI tools can often outperform a small team of juniors. However, relying only on senior talent is shortsighted. Companies must grow new talent, and that requires mentorship.

The best model, according to Martynas, is pairing senior domain experts who are also willing to coach with junior professionals who are eager to learn. Juniors often struggle to distinguish between accurate AI outputs and convincing but flawed ones. That’s what experienced experts can help them navigate. In such an environment, coaching becomes a critical skill for senior experts.

Despite some prejudices, juniors still create value, often even more than before. Martynas mentioned an example where interns, empowered by AI tools, delivered high-quality research and insights in just a week. This shows that with guidance, even less-experienced workers can have a meaningful impact.

At the same time, in some large corporations, certain seasoned professionals contribute little. They often have roles where domain expertise is kept, but output is minimal. For companies focused on long-term value, nurturing motivated juniors may prove more impactful than maintaining such veterans who are not engaged in their processes anymore.

AI in 2030: Expert predictions

Looking ahead, Martynas sees both opportunity and risk. As a father, he is curious and anxious about the world his children will enter. By the time his daughter finishes school, universities may look nothing like they do today (or might not exist at all).

His biggest concern isn’t mass unemployment but a potential loss of purpose. With AI writing, coding, and even creating art, people may question the value of their own contributions. This existential crisis could become one of the defining challenges of the next decade.

At the same time, AI may unleash entirely new services, industries, and creative possibilities that are hard to imagine today. Live, human-driven experiences, like concerts or conversations, will become even more treasured as digital content becomes abundant.

As Martynas noted, by 2030, people will live longer and create more personalized products (music, films, and even businesses) using AI not as a distant tool but as an everyday partner. 

At the end of this talk, Martynas joked that creating AI could have been the key purpose that people had on Earth. This legacy could outlive us and would stay here forever.

Whether it’s true or false, only time will tell.

But one thing is clear: when applied correctly, AI has enormous power to streamline both business and everyday tasks. 

Want to explore the AI space further and gain insights from industry leaders? Don’t miss the upcoming episodes of the Innovantage podcast, where Max Golikov will discuss the most pressing topics in the business and tech worlds with his guests.