Levels of Autonomy

The key to understanding how Autonomous AP works is to first understand the levels of autonomy with the help of the “Autonomous Index”. This index explains five active levels of autonomy. 


Level 0 (Manual Mode): In this level, all the invoices are manually approved. Thus from an automation point of view, this level has 0% autonomy. Hence, it is called the Manual Mode or Level 0 autonomous systems. Fully manual systems fall under this category.

Levels 1, 2, and 3 (Assist Mode): Systems that are semi-manual or need human intervention, for e.g., editing the required fields, are called Semi-Autonomous or Assisted systems. The level of automation in these systems can range from 20% to 60%. Here, the AP tools and applications assist AP processors with the data, defaults, workflow, and insights to make better decisions. Customers using Enterprise Resource Planning (ERP) applications typically fall in levels 1, 2, and 3. 

Level 4 (Autonomous Learning Mode): Autonomous learning systems are based on semantic understanding of the invoices. These systems use Machine Learning (ML) to learn and train themselves from past user actions. Autonomous AP fits at level 4 and helps achieve full autonomy in an average of 80% of the invoices.  upto 80% of the invoices. 

Level 5 (Autonomous Mode): Level 5 systems are known as Fully Autonomous systems. In such systems, the AP operator does not have to review or edit any part of the invoice. Autonomous Mode is the highest level with 100% autonomy. No technology has evolved to Level 5 or is fully autonomous yet.


The placement of AP in Autonomous Learning Mode is after considering the average of all the historical and live invoice findings. At the beginning of the automation, the system will fall under assisted mode whereas, with more user inputs and learning, it will move towards full autonomy. You can map the overall percentage of autonomy achieved by AP from your Analytics Dashboard. 

Where does Accounts Payable fit within the Levels of Automation

Just as AI transforms driving with autonomous vehicles, it also changes how invoices are processed at excessively high rates, even for non-electronic invoices. Both technologies work similarly. A Tesla car sees its surroundings by using computer vision. It can distinguish between people, stop signs and other vehicles to understand the current situation that the vehicle is in. Then, it uses AI to decide how to react. AppZen Autonomous AP reacts the same way to process and improve most invoices without requiring human intervention. 

Was this article helpful?
0 out of 0 found this helpful



Please sign in to leave a comment.