Machine Learning Vs AI

8th October 2018 | By Jan Cahill

For many, all this talk of robotics is a bit of a mystery.  Terms like machine learning, AI and process automation can cause some confusion,Digital Workforce - Administration Tasks so following a cuppa and a chat with our RPA Ninjas, Tom and Stewart, here are some of their thoughts and insights, find out more about the difference between AI and machine learning.

Machine Learning and AI

Machine Learning is a field of computer programming, where you give the computer data and it figures out the answer based on statistics.  For example, machine learning techniques are used to identify what type of dog is in a picture (check out this website!).  Every time it successfully gets an outcome, the computer learns and over time gets better with its answers. Self-driving cars also work on machine-based learning. According to our RPA Ninjas when you do those spam capture exercises where you identify all images with a street sign, or shop front etc. you’re actually helping to train a robot somewhere!

Another example is Microsoft’s Project Oxford.  This allows developers to tap into its resources using API (a programming interface) to build more intelligent applications. This project works on machine-based learning and can provide face recognition, speech processing, visual tools and language understanding intelligence.

Tip – Machine learning is only as good as the data you’ve trained the system on.  If you haven’t got enough varied data in the training the results will not be effective. It calculates probabilities and can give more weight to certain indicators.

AI (Artificial Intelligence) is a high-level concept for intelligent automation.  It’s the development of systems with the ability to do tasks which require human intelligence. True AI can think for itself in context, make decisions and have visual perception. You can train it to do many things, but there’s no way it tells you why it made that decision. It just makes it.

What is Process Automation

Robocloud provides its cloud-based software robots using Process Automation. It’s a logical, programmatic element which is process driven and based on binary decisions.  Processes are generated in programmatic steps and interact with software in the same way a human would.  We are able to see a full audit trail of why the robot has made a decision, as the software shows the decision trail all the way along, demonstrating what the robot has done and why.

Robocloud’s digital workforce can perform any manual task using process automation to save time and be more productive, for example:-

  • Process Incoming orders
  • Employee onboarding
  • Timesheet Automation
  • Supplier onboarding
  • Client onboarding
  • Statement Reconciliation
  • Account Reconciliation
  • System Integration

Process automation can also combine with the AI and Machine Learning techniques outlined above to provide a really compelling proposition – in our experience though most organisations will get the most benefit from automating the simple processes first to drive efficiency and data integrity benefits, which could then provide a platform to move on to more intelligent solutions.

For more information about the difference between AI and machine learning or Robocloud’s Process Automation Services call our team on 0800 046 8086 or contact us.