Generative AI is more than a technological leap – it’s a shift in how we create, collaborate, and solve problems. Turning complex AI capabilities into accessible tools is reshaping industries and redefining possibilities.
From automating creativity to enhancing decision-making, generative AI is driving change in how we live and work. It also empowers us to innovate like never before.
What is AI?
Simply put, AI is a machine’s ability to perform tasks that we, humans, find mentally challenging. AI mainly relies on machine learning (ML), which is how machines learn from data. Therefore, data is at the core of AI. Without quality data, AI cannot function effectively, as most of us working in the field know.
Analytical AI vs. Generative AI
In recent years, we can distinguish between analytical AI and generative AI. Analytical AI solves analytical tasks faster and more efficiently than humans. On the other hand, generative AI generates new data and content such as text, images, and sound.
One of the powerful aspects of generative AI is its ability to augment human tasks with simple natural language instructions (“prompts”). Additionally, the same models can be used across a wide range of tasks, making it highly versatile. In contrast, analytical AI is typically tailored by expert data scientists to perform specific tasks using complex methods.
The choice between analytical AI and generative AI depends on the specific problem at hand. Analytical AI excels in tasks that require data analysis, pattern recognition and decision-making based on historical data. For instance, it can be used to predict financial events by analyzing past trends and patterns.
On the other hand, generative AI, because of its versatility and ease of use, is highly accessible to individuals without the need for expertise in data science. This accessibility makes generative AI an excellent tool for innovation, allowing laymen to create and experiment with new ideas. For example, it can be used to generate creative content or simulate various scenarios.
We can also use analytical AI and generative AI together. In our delivery projects, we are using generative AI to alert us of relevant risks while correlating these alerts to the prediction of financial events using analytical AI. This combination allows us to leverage the strengths of both types of AI to achieve more comprehensive results.
Challenges and Solutions
At Wärtsilä, we have been developing AI solutions in production for several years. Our approach to AI solutions aims to maximize the value derived from data and AI technologies. In a business environment, the primary focus should be on solving business challenges in the most effective manner.
To achieve this, the first step is to assess whether AI is the right tool for the job. If it is, the next step is to integrate AI into a solution that includes IT development and change management aspects. This approach ensures that AI solutions are not only technically sound but also aligned with business objectives and operational needs.
Implementing such an approach requires certain capabilities in place. Over the years, we have invested in enriching our data assets, building modern data platforms, recruiting talents and establishing capable teams. This allows us to move quickly from idea to production.
Empowering Employees with Generative AI tools
When OpenAI introduced ChatGPT a couple of years ago, we quickly recognized and embraced the potential of generative AI. To maximize its benefits, we started piloting ready-made solutions like Microsoft 365 Copilot and GitHub Copilot to boost productivity, while also developing our in-house capabilities.
Alongside the use of these tools, we invested in practical training for all levels of the organization to raise awareness of the possibilities. These training sessions are often tailored to the specific needs of the target audience (such as top management, Salespeople, R&D engineers, etc.) to address the question of “what’s in it for me?”. During 2024, we have provided training to about a third of our total workforce, and there is still demand for more.
Our own in-house developed tool, WärtsiläGPT, provides various generative AI features to everyone in the company in a secure and compliant manner. Alongside a user-friendly web interface, we offer these features via an API, enabling easy integration with other systems and business processes.
Our approach combines the development of general functionalities with business case-specific solutions. This synergy allows us to leverage general features to support more specialized business solutions. Additionally, tailored functionalities created for specific use cases are made available to all users. For instance, beyond the chat feature, we offer enhanced capabilities for revising text, comparing documents, explaining terms, analyzing surveys, and more. Integrating company data and context adds additional value.
Significant Impact and Future Steps
The impact of adopting generative AI tools has been significant. We have seen steady growth in usage and an estimated worth of over 2 million € in work time savings during 2024. This figure excludes additional benefits such as improved quality and overall efficiency gains.
By making these functionalities available to all employees, we have transformed the perception of AI from an abstract concept to a practical tool that can be utilized in daily work. The adoption rate has been increasing rapidly, with a couple of thousand weekly users to date. Employees have been discovering innovative ways to work smarter and more efficiently.
The next step for us is to find ways to scale the innovations that individuals have been coming up with to the whole company. This will ensure that everyone can benefit from the advancements and continue to drive efficiency and innovation across Wärtsilä.
Global Hackathon: Showcasing Generative AI
One of the main challenges has been identifying new ways to solve business problems with AI. This is challenging because business people typically are unaware of how AI could be applied in their domain, and data scientists often don’t know enough about the problem and how to validate the usefulness of possible solutions.
An effective way for us to tackle this challenge has been to collaborate with our counterparts in the business via co-creation events such as hackathons. These hackathons bring together people with business and domain understanding, as well as developers and data professionals, to work intensively for three days to solve a business challenge. This helps bridge the gap between technical and non-technical people, create trust and accelerate the development of new solutions. Above all, the co-creation events bring excitement that is contagious throughout the organization, winning over even the most skeptical.
In November 2024, we culminated a year-long campaign with a global hackathon centered around the theme of generative AI. This event marked the pinnacle of our efforts, showcasing the innovative spirit within our organization.
Throughout the year, we encouraged our employees to explore and submit their ideas, resulting in an impressive 130 submissions. Nearly 2000 employees actively engaged in browsing and discussing these ideas, reflecting the widespread interest and enthusiasm for generative AI.
From the 130 submissions, we selected 7 standout ideas for the hackathon. A total of 58 participants from 16 countries came to our Helsinki headquarters. We divided them into teams, each tasked with developing a working Proof of Concept (PoC) that integrated various methods of generative AI, including text, voice, and image.
The winning team, “Speedy Mind-Readers,” clinched the top prize with their innovative AI-powered work reporting process. Their solution simplifies reporting into three steps: talk, review, and submit. By using generative AI to convert speech into text, engineers can verbally report their work while on the job, streamlining the entire process.
In the next few months, we will continue to develop the production solution. This will enhance efficiency and accuracy in work reporting across the company. At the same time, we will publish some of the generic functionalities developed in the project to all users in the company via WärtsiläGPT. This will make these advanced tools accessible to everyone.
This hackathon highlighted the potential of generative AI tools to drive innovation. It demonstrated how quickly these solutions can be implemented when the right co-creation setting and technical capabilities are in place. All selected ideas presented novel solutions to business challenges, leveraging the latest generative AI technologies.
The event showcased the power of empowering our people with advanced AI tools, leading to pioneering advancements and practical applications within our industry.
Here are some magical moments from the Generative AI Hackathon:

About the Author

Tal Katzav is the General Manager of the AI Centre of Excellence at Wärtsilä, based in Finland. With over 20 years of experience in Data Analytics, Machine Learning and AI, Tal leads a team of data scientists and machine learning engineers. Since 2017, the team has developed and productized more than 30 AI production-grade solutions across the enterprise. Tal is passionate about meeting new people and exploring innovative ways to solve business problems through technology. When he’s not busy with AI, you might find him trying to teach his computer to tell dad jokes. Because even machines need a sense of humor!
Add comment