Skip to main content

AI can support in every step

Artificial intelligence can be used in all parts of the pharmaceutical chain. That is the opinion of Christian Guttmann, global AI manager at the technology software company TietoEvry.

CHRISTIAN GUTTMANN was the global head of AI at Tieto, which since our podcast conversation in early 2019 has partnered with Evry to form the twice as large organization TietoEvry, with 24,000 employees. Guttmann is also active in two other roles – partly as a member of the Nordic AI Institute, partly at the Karolinska hospital where he supervises doctoral students in their research assignments. He notes that it is the capital-strong American market that controls investments in AI for the pharmaceutical sector.

– Many people are now looking at digital platforms that can increase the availability of their medicines, among other things by promoting the interaction between doctors, patients, and pharmaceutical companies, Guttmann says.

Increased availability with AI As the regulations regarding patient communication are extremely strict, artificial intelligence is being developed to increase the prescribing of certain drugs. But Christian Guttmann does not believe that it is necessarily a question of working in a legal and ethical gray area.

– It is rather a more advanced way of increasing accessibility. For example, you can get detailed information about what treatments other similar patients have received and how they have experienced it. This comes with great value since AI can analyze which drugs a patient is taking, as well as how well they work and interact. PatientsLikeMe is a good example of such a platform.

The goal is to build more services where updated research information can result in advice based on the latest scientific articles. The fact that both patients and care providers are becoming more educated is a great benefit, says Christian Guttmann, and adds that AI in healthcare has more areas of use than one might first think.

– AI is already present in every step of the drug process. It is being used in diagnostics, molecular analysis and to create new types of antibiotics with time savings of several years. Researchers who test manually in wet labs may be able to do 30-40 analyses a week. With AI and a database, you can test hundreds of millions of molecules and work much faster. Others use algorithms to produce representative experimental groups so that clinical trials can be carried out more quickly.

Amy books appointments

In addition, there is of course the possibility of having an AI create new, tailor-made medicines.

– We still have strict legislation that requires evidence in the form of clinical studies. But since AI quickly can perform a patient analysis of genes and microbiome, it is possible to see if there are patterns in a dataset. Christian Guttmann also practices what he preaches. When I send him an email with an interview proposal, it is his virtual assistant Amy who answers and sends a link to his calendar for booking the appointment.

– It works quite well. She can also make transcripts from my Teams meetings and summarize the action points that apply. It’s fun to test AI in a real-life situation to see this development up close.


"Researchers who test manually in wet labs may be able to do 30-40 analyses a week. With AI and a database, you can test hundreds of millions of molecules and work much faster."