• Written by Azhar Aliyar, Director

A purchasing manager submits a query to AsscherAi, asking for the total cost of a specific component across all current contracts. The system returns an answer, but the manager notices that one of the contracts is missing from the result. This is an example of a wrong answer, and it raises the question of whether to keep a human review step or remove it, in order to price the cost of such errors.

What each option genuinely commits you to

Removing the human review step commits you to relying solely on AsscherAi's algorithms to generate accurate answers. This means that any errors or inconsistencies in the data will be reflected in the answers provided, without any opportunity for human intervention or correction. On the other hand, keeping a human review step commits you to dedicating staff time and resources to reviewing and verifying the accuracy of AsscherAi's answers.

This is not a trivial consideration, as the time and resources required to review answers can be significant. However, it is also important to consider the potential consequences of incorrect answers, such as financial losses or reputational damage. By keeping a human review step, organisations can help to mitigate these risks and ensure that the answers provided by AsscherAi are accurate and reliable.

The case where the other one is right

There are certainly cases where removing the human review step is the right decision. For example, in situations where the data is relatively simple and straightforward, and the queries are routine and well-defined, AsscherAi may be able to provide accurate answers without the need for human review. In such cases, removing the human review step can help to streamline processes and improve efficiency.

On the other hand, there are also cases where keeping a human review step is essential. For example, in situations where the data is complex or sensitive, or where the queries are nuanced or open-ended, human review may be necessary to ensure that the answers provided are accurate and reliable. In such cases, removing the human review step could lead to errors or inconsistencies that could have serious consequences.

The cost that only shows up later

One of the challenges of deciding whether to keep a human review step or remove it is that the costs of removing it may not be immediately apparent. For example, if AsscherAi provides an incorrect answer that is not caught by a human reviewer, the consequences may not be apparent until much later, when the error has already caused financial losses or reputational damage.

In order to fully understand the costs of removing the human review step, organisations need to consider the potential long-term consequences of errors or inconsistencies in the answers provided by AsscherAi. This may involve conducting a thorough risk assessment, and considering the potential impact of errors on the organisation's finances, reputation, and operations. For more information on how to conduct a risk assessment, you can visit our website.

A way to decide without a six month evaluation

While a six month evaluation may provide valuable insights into the effectiveness of AsscherAi with and without a human review step, it is not always practical or necessary. Instead, organisations can use a variety of other methods to decide whether to keep a human review step or remove it. For example, they can conduct a pilot study, or gather feedback from users and stakeholders.

Another approach is to use a decision tree, which can help to identify the key factors that should be considered when deciding whether to keep a human review step or remove it. This can include factors such as the complexity of the data, the nuance of the queries, and the potential consequences of errors. By considering these factors, organisations can make a more informed decision about whether to keep a human review step or remove it, without needing to conduct a lengthy evaluation.

What this does not do

This article does not provide a definitive answer to the question of whether to keep a human review step or remove it. Instead, it highlights the importance of considering the potential costs and consequences of each option, and provides guidance on how to make a more informed decision. It also does not address the potential benefits of using AsscherAi in conjunction with other systems or tools, such as data analytics software or machine learning algorithms.

For organisations that are considering using AsscherAi, it is essential to carefully evaluate the potential benefits and risks, and to consider the specific needs and requirements of their organisation. If you have any questions or would like to discuss your specific needs, you can contact us through our contact page. By taking a thoughtful and informed approach, organisations can get the most out of AsscherAi, and ensure that they are using it in a way that meets their needs and supports their goals.