Artificial intelligence has developed rapidly, especially its applications and utilisation possibilities. Currently, the cutting-edge opportunities of AI research are focused on diverse AI and agents. Those who have watched the development from the sidelines and don’t want to be early adopters are now wondering what benefits AI could bring to them or their organisation. It’s hard to see the opportunities that the latest AI innovations bring, or they seem far away.
Microsoft’s Copilot and OpenAI’s ChatGPT already offer AI-powered tools that can be easily integrated with your organisation’s data and tailored to your needs. These tools can also perform tasks on behalf of the user or with the user’s approval. Let me concretise this with an example: When analysing source material with ChatGPT, you can ask AI to record observations in a memo program that the user uses.
Everyday challenges and how AI can help
The examples may sound simple and trivial at first glance. However, the everyday life of a modern knowledge worker is full of jumping between different tools, and there are hardly any uniform experiences. Information is scattered in different places, some are restricted by user rights and some are public. When these repetitive, albeit simple, tasks are automated, work becomes more efficient. Meaningless and repetitive work is reduced.
Recently, the use of AI has been considered in various tools related to the terms and conditions of employment and instructions of personnel. The use case is clear: what if an employee could get an immediate response about their vacation days, standby compensation, occupational health conditions or teleworking practices? This information is usually recorded on the workplace intranet or found in the information of an HR system, but what is new is that it is easily accessible to the employee.
Some readers may already snort that artificial intelligence and search engines are getting confused again, and I can’t completely deny that. Over the years, organisations have faced the challenge of how to effectively communicate guidelines, rules, and policies to staff, especially when sources are diverse, such as legislation, collective agreements, local agreements, and work instructions. But there’s more behind it: AI can provide information such as a person’s years of employment, employment contract, and department that influences the response. The answer is therefore tailored to the questioner. At the same time, the answer includes references to source materials, so that the questioner can easily access the original source of the information, which isn’t necessarily a single document but can be, for example, an HR system.
This is just the beginning. The use of an AI application doesn’t end with a chat response but can continue from there and do things for the user, such as notifying the person of sick leave or ordering new devices for the person’s use.
And now I come to that steak: All of the above is already possible and relatively easy. It isn’t state-of-the-art science, but AI-assisted solutions that are already coming to a practical level.
Agents and AI now
Agents are the keyword in the AI hype of 2025. In practice, an agent means that artificial intelligence isn’t only used as an interlocutor, but can also be used to do sensible things that genuinely help a person or organisation. While there are still a lot of unfulfilled promises here, there are also ready-made technical capabilities that can be deployed.
Over the past year, an established rule of thumb has emerged: it is usually not worthwhile for an organisation to build new AI capabilities from scratch if there are already ready-made features available from AI suppliers for the same need. The development of these functionalities is progressing rapidly and new solutions are constantly entering the market. In other words, you should first take advantage of ready-made and easy-to-deploy solutions before making large investments in your own construction.
Low-code platforms like Microsoft Power Platform meet this need perfectly. Whereas in the past it took significant technical investments to create a chatbot based on organisational data, today the first working version can be made with low-code in an hour. The same goes for agents. If you want to try out how it works to automatically assign messages to a shared RFQ mailbox based on content analysis to suitable managers, you don’t need to start a large IT project, but reserve an afternoon.
However, it should be borne in mind that tools are able to quickly provide fairly simple solutions. When an organisation is multilingual, it is subject to legislation in different countries, and even the source material – the famous data – is what happens, there will be no high-quality implementation in the afternoon. But in any case, understanding increases, and when people see concrete opportunities, they begin to explore the potential of utilisation in other contexts as well, and thus the possibility of finding the right use cases increases.
There are plenty of organisations in Finland that use the services of Microsoft or other vendors and ready-made AI capabilities. It is really easy to test the benefits of these and find solutions that genuinely benefit everyday life. AI isn’t only the exclusive preserve of global giants, but it can be utilised here and now, even in small tasks.
From small streams, one percent at a time, organisational efficiency and job satisfaction improve and benefits accumulate. This is already possible right now.
Want to read more about low-code?