Originally published in Bonfire business media.
Artificial intelligence constantly crosses our attention with either success or horror stories. On one hand, we hear AI will free people from routine tasks, on the other, we are told AI will take away our jobs. Companies showcasing different pilot projects often feed into the AI hype.
Despite the media attention, company leadership and especially boards of directors should stop and think about their business: How can we, instead of participating in the hype, focus on the essential and systematically improve our business with these new technologies?
Although AI can improve your business significantly, the development and on-boarding requires similar efforts as any other new technology. You shouldn’t expect miracles such as productivity leaps, unless you do your groundwork thoroughly.
On-boarding AI starts with analyzing the current state
Any development process starts with assessing what you have in place today. Companies should map out and seek to understand the state of AI in different business functions, especially in terms of the company’s products and services. Leveraging AI is not limited to the core business, but also any supporting activity can be improved with AI.
The maturity level for on-boarding AI should be investigated across the organization: in customer interaction, IT, finance, legal processes, marketing, sales and business development and production.
If the company is already using AI, it is important to understand how the AI system learns. It is fundamentally different to build a system that constantly learns from accumulating data and feedback, than using a pre-trained model that won’t improve as the new data flows in.
Think about Google’s search algorithms that improve and optimize every time a user chooses a certain result based on a certain keyword.
The existing infrastructure integration and requirements, such as how data is collected and stored and if the company has RPA in use, all have an impact on the AI development. The development is also affected by general digital capabilities and ability to access different types of data.
Companies should create their data management from an AI point of view, making sure that data is not biased and data privacy is respected.
AI maturity levels can be categorized as follows:
- No activities with AI
- Use cases identified
- Pilots running / run
- Commercial use of AI
- Leader in using AI
In Finland Ramboll has been piloting machine learning that improves the quality of water. Running pilots and building proof-of-concept solutions is an effective way for validating if the model is working properly and the existing data is sufficient for the purpose. In the next phase the improved model can be deployed in commercial use.
Prioritizing and deciding on development areas
After the current state evaluation, it is important to think about what to prioritize. The AI maturity levels presented above help the company and its leadership and board to get a holistic understanding of applying AI at the company.
Furthermore, the company needs to decide if it seeks to develop tailored solutions or prefer to use ready-made AI products and services. Once the full picture has been drawn, it is easier to establish requirements for both internal and external stakeholders.
Business transformation is always a strategic decision. When applying AI to the business it is crucial to understand which products and services are most business critical and have the most potential for leveraging AI. Strategic development is tied to how the leadership and personnel view the use of AI in the company.
Systemic improvements lead to competitive advantage
Once the analysis of the current state has been systematically completed, and the results documented and updated regularly, the unnecessary hype can be replaced by constantly improving business with AI. A partner with expertise in building AI solutions can help with this process.
Ideally the development of AI is not a separate process, but an integrated part of the company regular business activities. This way AI starts to create competitive advantage. As end result, the core processes of the company generate data for the AI to learn from, and the human experts’ feedback keeps improving the model to become even better.
The selected services and products are being improved with AI to gain competitive edge, and the company’s leadership and personnel are committed to the AI-driven strategy. In addition, the company has identified possible areas of further development and offers adequate training for the personnel in recognizing them.
In case it’s needed, the company can procure expertise from external AI specialists. A good example is Finnair where we Silo AI provided experts to the building a solution that improves the situational awareness at the flight operations control with machine learning. The predictions of the model constantly get more accurate as new data comes in. In the next phase the company will take the model into commercial use as part of a larger technological update.
It is not beneficial to stay with the hype or fear that you’re already too late. The AI transformation is coming slowly but surely. By systematically improving your business processes, it is possible to start learning and improving faster than your competitors. You don’t have to do everything by yourself, but rather leverage already existing AI solutions or extend your own AI expertise with expert consultants.
Come and listen to Tero Ojanperä (Silo.AI), Catherine Havasi (Snowcap AI) and Colin Shearer (Houston analytics) take the stage at SHIFT Business Festival in Turku on August 30th.
About
Join the 5000+ subscribers who read the Silo AI monthly newsletter to be among the first to hear about the latest insights, articles, podcast episodes, webinars, and more.
