These days, we hear about many hot IT concepts and it can be easy to confuse them.

[Read the main article, an interview with If’s head of robotics Asko Mustonen:
• His new robots will do 2 million tasks this year
The other side articles:
• Cancellation robot makes changes in 5–6 systems
• RPA is among the simplest IT solutions]

In moving from monolithic legacy systems, many financial companies seek to combine different best-of-breed parts, which they connect through standardized application programming interfaces, APIs.

Increasingly, these systems are provided as services across the internet, “in the cloud”. While these cloud and API integrations can seem like lightweight in comparison to old systems, they are still a more ambitious route than a robotic processing implementation.

Another often-heard term today is artificial intelligence, AI, including machine learning. This applies to the improving capability in software to interpret data, and handle it accordingly, depending on its context. This can support translation of speech into text, for example, or make it possible to feed a data flow such as accounting with pdf files of unspecified structures.

Compared with these solutions, robotic process automation typically limits itself to data which is already structured enough to be handled without variation or ambiguity. This makes a radical difference to the need for software complexity.

Read the main article, an interview with If’s head of robotics Asko Mustonen:
• His new robots will do 2 million tasks this year
The other side articles:
• Cancellation robot makes changes in 5–6 systems
• RPA is among the simplest IT solutions