Arago’s AI-based Business Process Automation – A Discussion with René Büst
René Büst, Director of Technology Research at Arago, was featured with his views on the impact of artificial intelligence (AI) on businesses as well as the possibilities it has to offer in two podcasts in November. He was interviewed by Rozee, a podcast on AI transformation in businesses, as well as by dotmagazine, the online publication of eco – an association of European companies in the internet economy.
Büst prominently presents Arago’s general AI solution HIRO™. Most AIs available for consumers would be so-called narrow AIs, such as the technologies behind Spotify, Netflix or Pandora. They would be able to recommend music and movies based on one’s own taste and social network but are not capable to solve other issues. In contrast, Arago’s general AI HIRO™ can take over processes in any given scenario or company. He also argues that Spotify and Netflix have been some of the first enterprises to utilize cloud computing as well as AI. He calls their approach the “magic triangle”: The companies use the cloud for their products, apply AI technologies to better understand customers and deliver more customized products to their users. In his opinion, AI is a logical next development to cloud technology. The big cloud services of Google, Amazon and Microsoft would have realized this as well.
In order to improve AIs and feed them with data, the main goal of developers should be to reach higher engagement rates with the users. This would enable companies to offer individually customized products to clients. While most consumers still use a keyboard to request information, they will increasingly speak with their devices. Microsoft’s Cortana and Amazon’s Alexa would be prominent examples. Büst is convinced this development will give companies the ability to offer 24/7 customer service with AI bots. In particular, small companies, which are not able to afford call centers, could benefit from such AI systems.
Büst considers AI a game changer for enterprises in established industries as well. At the moment, their business models would be heavily challenged by online platform companies. First, tech giants such as Google or Alibaba would have immense cash available and could easily spend money to branch out into areas outside their core business models. Second, they would hijack the customer relationships of traditional companies by forcing them to offer products through their platforms. As a result, established brands find themselves as one amongst many in the vast platform pool. At the same time, the direct customer interaction takes place between the platform and the consumer. AI solutions would offer established enterprises the possibility to automate processes, save money as well as time and to use the freed-up resources to innovate. As such, they are able to become disruptors themselves. Büst also argues that companies should not just consider adopting a single technology. Instead, they should reflect how to improve their business with a variety of technologies such as AI, the cloud and blockchain. A natural starting point for AI automation would be a company’s IT infrastructure since all business data would pass through the IT level. In order to adopt AI solutions within an enterprise, Büst believes that companies should rather possess skills in knowledge management than in computer science. A special “company AI manager” would, therefore, be unnecessary. Companies should remember that a giant data pool alone could not be leveraged by an AI. For an AI to make sense of the data, the enterprises would have to think about ways to teach it with the company’s relevant processes. The businesses adopting an AI should ask their employees how they are running processes. This could be, for example, through interviews or feedback the employees give to the AI about its proposed solutions. As a result, AI would not only give companies an edge in competition but also insures them against the loss of knowledge if an employee leaves the enterprise.
Büst further argues that AIs could be considered “buddies” for employees in the context of network security. AI capabilities would be perfect to analyze data as well as log files and to spot anomalies. In this way, they would not replace humans in IT security but assist them. Overall, he foresees repetitive jobs to disappear, while employees with individual skills would have a bright future in the current industrial revolution. Gut feeling and empathy would also become increasingly relevant skills. At the same time, he does not consider the potential inability of machines to make the right ethical decisions to be an issue. In the end, humans would still be the main influence on an AI’s behavior.
Click here to listen to the podcasts: