• Written by Sonali Subhadarsini, Product Manager

On a typical Monday morning, the executive leadership team gathers to discuss the previous week's performance, and the shift supervisor's maintenance log is just one of the many reports they need to review, but what they really want to know is how the latest sales figures are impacting the company's overall financial health, and whether the current headcount is sufficient to meet the growing demand.

Enforcing permission at retrieval

When implementing a system like AsscherAi, it's tempting to try to instruct the model on what data it can and cannot access, but this approach often falls short, as the model may not always understand the nuances of the organisation's permission structure, and it's not uncommon for the category manager to have access to certain supplier information that the shift supervisor does not, and enforcing these permissions at the point of data retrieval is a more effective way to ensure that sensitive information is only accessible to those who need it.

The problem with instructing the model is that it can be difficult to anticipate all the possible scenarios in which data might be accessed, and it's easy to overlook certain edge cases, such as when a new employee is hired and needs to access certain data to perform their job, but has not yet been added to the relevant permission groups, and by enforcing permissions at retrieval, the system can ensure that the correct access controls are applied in real-time, regardless of how the data is being accessed.

Auditability

Auditability is a critical component of any enterprise AI governance system, as it allows the organisation to track what questions are being asked, what data is being returned, and on what data the responses are based, and this information can be used to identify potential security risks, or to troubleshoot issues with the system, and the category manager can use this information to ensure that the supplier data they are accessing is accurate and up-to-date.

Most people get wrong the idea that auditability is just about tracking who accessed what data, and when, but it's actually about understanding the entire workflow, from the question being asked, to the data being returned, and how that data is being used to inform business decisions, and by having a clear audit trail, the organisation can ensure that the system is being used in a way that is transparent, and accountable.

Executive leadership data

The data that executive leadership relies on is typically consolidated financials, operational KPIs, sales performance, and headcount, and this data is often scattered across different systems, and departments, and the shift supervisor may have access to certain operational data, but not the financials, and the category manager may have access to sales data, but not the headcount information, and by bringing all this data together in a single system, AsscherAi can provide a more complete picture of the organisation's performance.

This data is critical to informing business decisions, and by having it all in one place, the executive leadership team can get a better understanding of how different parts of the organisation are impacting each other, and make more informed decisions, and the executive leadership team can use this data to identify areas where the organisation needs to improve, and develop strategies to address these issues.

What it changes

One of the biggest benefits of AsscherAi is that it can help to answer cross-functional questions that nobody can answer in the room, such as how a change in sales strategy might impact the company's financial performance, or how a new operational process might affect headcount, and by providing a single source of truth for all this data, AsscherAi can help to break down silos, and improve communication between different departments.

Each function often reports its own version of a figure, and this can lead to confusion, and inconsistencies, but by having a single system that can provide a unified view of the data, the organisation can ensure that everyone is working with the same information, and this can help to reduce errors, and improve decision-making, and the category manager can use this information to ensure that the supplier data they are accessing is consistent with the sales data, and the shift supervisor can use it to ensure that the operational data they are accessing is consistent with the financials.

Measuring success

Measuring the success of an enterprise AI governance system like AsscherAi can be challenging, but there are a few key metrics that can be used, such as decision latency, which is the time it takes for the organisation to make a decision, and actions of the form come back with numbers, which is the ability of the system to provide actionable insights, and definition conflicts resolved, which is the ability of the system to resolve inconsistencies in the data, and by tracking these metrics, the organisation can get a better understanding of how the system is performing, and where it needs to be improved.

By using these metrics, the organisation can ensure that the system is having a real impact on the business, and that it's not just a technology for technology's sake, and the AsscherAi team can work with the organisation to develop a customised metrics programme that meets their specific needs, and provides them with the insights they need to drive business success.

What it does not do

AsscherAi is not a replacement for human judgment, and it's not a system that can be used to make decisions without any human oversight, and it's also not a system that can be used to access data that is not already available to the organisation, and it's not a system that can provide insights that are not based on the data that is available, and by understanding what the system can and cannot do, the organisation can ensure that they are using it in a way that is effective, and responsible.