Understanding and application possibility of artificial intelligence.


Keep your finger on the pulse with RPA – Robotic Process Automation and make your company future-proof.


Very general term for “artificial intelligence” (abbreviated to “AI”), which describes the imitation of human intelligence by machines or automata. The term was coined at the historic 1956 Dartmouth Conference, considered the “birth of AI.” (Source: “Basic Course in Artificial Intelligence” by W. Ertel)

Science of describing, analyzing and using data (or more precisely multidimensional data arrays = tensors). Due to widespread storage capabilities and digitization, Data Science is taking on an increasingly important role today and has become indispensable. In particular, machine learning is used in the analysis and use of data. (Source: “Data Science for Business Intelligence and Business Analytics” (Lederer and Schmid, 2020)

Deep Learning describes machine learning by means of artificial neural networks, which comprise several layers (deep layers). The large number of layers is intended to make even difficult tasks solvable for machines.

Abbreviation for “Generative Pre-trained Transformer 3”. A product of the “Open AI” initiative in which Elon Musk, among others, was involved. At its core, it is an autoregressive machine learning model that has great strength in Natural Language Processing (NLP). (Source:

A front-end for TensorFlow written in the Python programming language. Developed by Francois Chollet. (

A machine learning model that specifically mimics neural networks of sentient beings with their “percolating” understanding mechanisms. The so-called backpropagation algorithm is often used here.

Logic is the basis of modern computer science, because not only numbers but also symbols can be processed. The first viable logic systems were presented at the historic 1956 Dartmouth Conference.

Machine Learning represents exactly one form of Weak AI, in which classical statistical methods have been or are being further developed (primary example: linear regression). Algorithms with corresponding parameters are an integral part of this.

Abbreviation for “Natural Language Processing”. This refers to all types of machine learning that deals with understanding, analyzing, and processing natural language of living beings (especially humans).

Very popular programming language for all types of machine learning. (

While Artificial Intelligence captures “thought processes”, the adjective “robotic” focuses on the artificial imitation of physical actions. These actions can refer either specifically to actions in software/computer operation (see also “RPA”) or more generally to all kinds of physical actions outside the operation of a computer. For the latter, of course, you need a physical robot.

Machine learning that works on the principle of “trial & error” together with rewards for success (“learning like a toddler”). It is especially common for physical robots that need to find their way in unfamiliar environments.

Artificial intelligence that acts on an equal footing with humans and is completely equal to them in all aspects. Currently, this is only a subject of fiction (especially science fiction), because the (technical) realization (despite many and historically far-reaching attempts) is very complex and difficult.

Artificial intelligence that covers only certain tasks that require human intelligence. By focusing on a specific aspect, greater precision and efficiency can be achieved compared to human agents. However, only in the clearly defined area of responsibility.

Machine learning with or without “pattern solutions”, both have advantages and disadvantages. While Unsupervised Learning is mainly used for generating ideas, Supervised Learning is mainly used for problems with large data sets and the focus here is on precision in task solving.

A software library written in the Python programming language for Machine Learning with a particular focus on Deep Neural Networks. Initiated by Google, now available as open source. The name “TensorFlow” refers to the processing of multidimensional data arrays (the basis of Data Science), which are also called “tensors”. (

Abbreviation for “Tensor Processing Unit”, a computer chip suitable for handling multi-dimensional data arrays (aka “tensors”).


Automation Anywhere is an American global software company that develops software for robotic process automation. Founded in 2003, the company is headquartered in San Jose, Calif. (Source:

Supervised robots work side-by-side with humans on their desktops, acting like personal assistants to help with daily tasks. Employees access the wizard, schedule it, and execute automations. (

Blue Prism is the trading name of Blue Prism Group plc, a British multinational software company pioneering and providing robotic process automation software for enterprises. (Source:

“Process management, also known as business process management, is concerned with the identification, design, documentation, implementation, control and optimization of business processes.” (Source:

“A business process is a sequence of activities for the fulfillment of an operational task. From one or more inputs (input), a customer-beneficial output (result) is generated, i.e., products are created or services are provided through information and material transformation.” (Source:

Hyperautomation is a business-driven, disciplined approach that enables organizations to quickly identify, review and automate as many business and IT processes as possible. Hyperautomation involves the coordinated use of multiple technologies, tools or platforms, such as Artificial Intelligence (AI), Machine Learning, Event-driven Software Architecture, Robotic Process Automation (RPA), Business Process Management (BPM) and Intelligent Business Process Management Suites (iBPMS), Integration Platform as a Service (iPaaS), low-code/no-code tools, packaged software and other types of decision, process and task automation tools. (

OpenRPA is the first completely free, open source platform for enterprise robotic process automation. (Source:

Optical character recognition, or OCR for short, is the use of technology to recognize printed or handwritten text characters within digitized documents such as a scanned paper document. The basic approach of OCR is to examine the text of a document and translate the characters into a code that can be used for data processing. (Source:

Process mining is defined as an analytical discipline for discovering, monitoring, and improving processes as they actually are, not as you think they might be. (Source:

“Robotic Process Automation is an approach to process automation in which repetitive, manual, time-consuming, or error-prone activities are learned and automated by so-called software robots.” (Source:

Screen scraping is the process of collecting screen display data from one application and translating it so that another application can display it. This is usually done to capture data from an older application so that it can be viewed through a more modern user interface. (Source:

Task mining is a service for detecting automation opportunities. It collects employee desktop data, consisting of screenshots and log data on each user action (mouse click, keystroke), and then runs a machine learning model to analyze the data and suggest a list of processes with high automation potential. (Source:

“UiPath is a US-based software company founded in Romania and headquartered in New York that operates a robotic process automation platform for business customers.” (Source:

They work independently in the background and complete task-intensive, long-running processes. Unattended robots report to staff for validation or questions or exceptions. (Source:


“A programming interface (API for short), is a part of a program that is made available by a software system to other programs for interfacing with the system.” (Source:

“Cloud computing describes a model that provides shared computing resources as a service, such as servers, data storage, or applications, on demand – usually over the Internet and independent of devices – in a timely manner and with little effort, and charges based on usage.” (Source:

“An interface refers to a transition point between different components of an IT system through which data exchange or data processing is realized. These can be human-computer interfaces or computer-computer interfaces.” (Source:

“On-premises or on-prem (on-site or local) refers to a usage and licensing model for server-based computer programs (software).” (Source:

“Off-premises or off-prem (not on-site or global) refers to a usage model for cloud-based software.” (Source:

Sources, unless otherwise noted: “Artificial Intelligence: A modern approach” by S. Russell and P. Norvig (published by Pearson) and “Grundkurs Künstliche Intelligenz” by W. Ertel (published by Springer Vieweg). Date?

You are still undecided and want to learn more first? Then let’s have a digital coffee together and you can ask everything, without any pressure. Just write Tobias via the coffee icon.