Why "artificial intelligence" will now change the competition
What is currently new in the use of artificial intelligence?
Sinan Tankaz: The use of AI in business has so far been characterized by analytical applications. A generation with completely new capabilities will now greatly expand the possible applications - the so-called generative AI models. One of the best-known examples is the "Generative Pre-trained Transformer", known to many as Chat GPT from OpenAI. Generative AI algorithms use existing data - videos, images or even computer code - and create new content from it. This creative element differs from the pure analysis function.
What does this mean for companies in concrete terms?
Sinan Tankaz: Now is an important time for companies to decide on the use of AI. The reason: when used correctly, the new applications of business AI are a lever that fundamentally improves competitiveness. Companies that get their act together succeed in using generative AI to develop very strong market advantages and leave their competitors, who have hesitated for too long, behind. CANCOM Austria has already been using AI for its customers in productive environments for 6 years and, among other things, ensures that output is maintained by preventing outsourcing.
What does this look like in practice?
Sinan Tankaz: Customer service call centers in all industries require enormous cost and personnel input. We have developed an AI-based assistance system for customer service that automates all repetitive work in the first step. The second step involves a leap in quality that goes far beyond the classic use of AI: the AI-based recommender independently recognizes and reports the ideal time to contact the customer, the preferred communication channel and which offer makes the most sense. For inquiries in the self-service portal, the system assists with product selection. This enables operators to provide a fast and targeted service that puts the traditional call center experience in the shade. The call center team, relieved of standard inquiries, now takes care of the difficult consultations with the time required. This type of AI strategy is particularly important for traditional companies such as banks and insurance companies. This makes it possible to reach special target groups such as "digital natives" or the rapidly increasing number of older people as a result of demographic change. In recent years, a provider such as Amazon has demonstrated the dynamic effect that a differentiated level of quality in customer service can have on a company's competitive position. When implementing AI, CANCOM Austria relies on the use of graph technology. Similar profiles with similar behavior are merged, i.e. dynamic data is generated and profiles are created.
How can this AI lever be used in industry?
Sinan Tankaz: We use machine learning in the manufacturing industry. Here, too, it is possible to strengthen the market position of customers with leaps in quality. Here's an example: the experts at CANCOM Austria developed a customized business AI for a manufacturer of large industrial transport bags - so-called "big bags": the aim was to ensure the customer's ambitious performance promise that the company's "big bags" would last 100 percent reliably even under maximum stress. This is important because the contents of the goods in such transport bags usually exceed the value of the carrier bags by a factor of a thousand. The new AI application is now working very successfully in production monitoring at the manufacturer. Sophisticated sensor technology provides seamless data that is evaluated using artificial intelligence. The algorithms learn to detect the smallest deviations in product quality and play to the strength of machines to tirelessly monitor 24 hours a day, 7 days a week. The result: the manufacturer was able to keep its performance promise and thus significantly increase the prices for the new product line of zero-defect big bags - the earnings per unit are now around four times higher.
AI often works with sensitive data - what is your experience here?
Sinan Tankaz: With our AI applications, for example, we are working on making everyday life easier for doctors with computer-assisted processes. CANCOM Austria works locally with universities and institutes to ensure quality in the development of AI in this sensitive area with patient data. The benefits of AI applications in medicine are also extremely high: for example, our AI engine scans doctor's letters and prescriptions and independently collects data from them. On this basis, the application provides information on medications, active ingredients and disease progression. This means that not only can medication be used optimally, but interactions can also be avoided. Doctors are relieved of precisely those areas that the machine can do faster and more reliably than they can. It goes without saying that the data processing of sensitive customer data in AI-based applications must be secured.
Speaking of security: what should always be taken into account with AI systems?
Sinan Tankaz: The use of AI models is based on trust. However, as AI systems work with neural networks and complex decision trees, it is more difficult to understand the decisions than with conventional software - the so-called black box problem. However, transparency is imperative when dealing with sensitive data in business. CANCOM Austria therefore relies strictly on trustworthy, open and explainable AI models. With these "explainable AI standards", decision transparency is defined, i.e. the question "How is it decided whether results are right or wrong?" and the comparison of identical questions is carried out in feedback loops. This follows a central motif of the European Union's Artificial Intelligence Act (AI Act), which aims to regulate AI in all areas of life for the first time worldwide.
What does the future look like?
Sinan Tankaz: Generative AI models will help to overcome the most important challenges currently facing companies. Success in practice is always about organizing the division of labour between humans and machines in the best possible way. The reason: what is particularly easy for humans is difficult for AI systems and vice versa. Software, for example, can tirelessly sift through and compare tens of thousands of pieces of data - a task that quickly tires people out. However, when it comes to listening carefully to a person and empathically grasping the subtleties, the machine cannot keep up. The combination of man and machine is therefore the model for success.
What is your conclusion?
Sinan Tankaz: Companies should now seize the opportunity to make use of the new possibilities offered by business AI. Automation with analytical AI is always a promising way to save costs, reduce the workload on employees and raise quality levels. However, generative AI algorithms can do much more. My recommendation to management: benefit from our practical experience and discover that artificial intelligence delivers tangible competitive advantages.