Full automation: a vision of the future

A timelapse photo of a port city with neon streaks representing movement | AI tools Unit4

Co-written by Rafael Artacho, product director at Unit4.

There’s no need to fear the machine. It offers insight that no human effort can replicate in a similar time window. AI tools aren’t going anywhere soon, maybe it’s best to examine the technology closer and divine the benefits it can provide to your manual processes.

Artificial intelligence tools continue to mature, and the rapid rise of generative AI has prompted near hysteria about its potential to replace everything from Google Search, to human effort. At Unit4, we have been working with AI and other innovative technologies for many years to transform the ERP system. Our digital assistant, Wanda, became the first ERP assistant using AI to automate tasks like filling in employee holiday or absence forms, or prompting manager approvals, significantly reducing the burden on users to interact directly with ERP.

There is some discussion about AI automating up to 50 percent of existing manual processes, which could mean AI tools may reach a point where there is little need for users to input data. For anyone who has worked with previous generations of enterprise software, such a vision sounds very enticing, but what would it really look like and how feasible is it?

What could it look like?

The Unit4 platform, from ERP to HCM and financials, is already being worked on to redefine the user experience. Machine learning models and existing software, like ChatGPT, will help implement smart, context-aware functionality that will support users wherever they are regardless of the platform, device or communication channel. The vision of the future is clear: data will be pre-populated, decision-making will become more informed, anomaly detection can be enhanced and risk management transforms into something a lot more proactive.

These developments hold tremendous potential to drive efficiency, accuracy and agility in financial processes, benefiting both businesses and individuals.

How realistic is it?

The AI tools are maturing rapidly, so there is a very real prospect they could dramatically improve the user experience in the coming years. We are actively collaborating with Microsoft and engaging in continuous development to enhance the software on hand.

However, it’s important to acknowledge the challenges and considerations that accompany these advancements.

Consider data privacy. The importance of data privacy, security and ethical considerations in deploying AI technology is something Unit4 recognizes. By strictly adhering to GDPR and monitoring evolving statutory requirements, we prioritize the protection of sensitive information and maintain the trust of our users. Furthermore, a regular auditing process has been implemented. The new process will work towards strengthening the system’s compliance with security measures, ensuring a responsible and secure AI-driven deployment.

Building trust with users is also crucial to the success of an AI-integrated system. Users need to feel confident that the system they’re interacting with is accurate and efficient. Unit4 uses AI models that are transparent and understandable, explaining to users how decisions are made. There is a conscious effort to eliminate ‘black box’ scenarios where outputs are generated without clear indication of how they were derived. To strengthen the trust factor, it’s imperative to prioritize a data approach that isn’t siloed but offers a global view, enabling the ability to identify common patterns and trends across the users.

Of course, we should not get carried away with the idea that AI removes the need for human oversight. There will always be a need for human involvement. Although we may no longer have manual processes in the future, we will still need people to check the AI’s decision making, ensuring that AI systems remain accountable and that they continue to serve the human-centric values of the organization.

Embrace AI now to avoid being left behind

It will be critical to ensure the use of AI accurately reflects what the user needs and wants from automation. This can be addressed by minimizing how much manual input is required from users, but we also have to balance that against the system being perceived to be interfering. That’s why it’s best to lean heavily on ethnographic research methods and user-centric design, to inform how the AI is being adopted so that data can be integrated into the user experience.

The Unit4 Experience Design and Automation team places tremendous value on this research, as it presents an opportunity to study users within their natural working conditions. By embedding in real-world environments, we gain the opportunity to obtain a comprehensive understanding of how our ERP system functions in day-to-day situations. This in-depth insight enables the creation of more targeted, efficient and user-centric solutions.

Unit4 had the chance to study a non-profit customer based in the US. This field trip allowed us to uncover user requirements that were difficult to express or entirely unknown to the users. For instance, areas such as ‘Expenses’ suffer from low screen resolution. By carefully observing user interactions,the areas where users needed additional support could be pinpointed , helping to refine the training and assistance processes. Opportunities were found to increase efficiencies within the ERP system by automating manual tasks, such as replacing manual payments with remittance. We have prioritized these discoveries and incorporated them into an actionable list for future implementation.

Our mission is to meet users in their context. This is the reason we approach user experiences from a holistic perspective. To craft the best user experiences, you need to look beyond the limitations of an ERP and consider a user’s whole ecosystem and every interaction faced to complete a task, regardless of system, platform, device or communication channel. We believe in a future where static, traditional user interfaces evolve into more dynamic, automated, end-to-end user experiences that are context-aware and adaptable. These experiences are intelligent and tailored to individual users’ specific circumstances and tasks, enabling seamless interactions. The primary objective is to increase users’ efficiency and satisfaction by significantly decreasing errors and eliminating manual, repetitive tasks.

Ultimately, the goal is to deliver accurate AI models that meet users’ needs. While there is still work ahead, we are committed to curating and actively incorporating user feedback to increase the quality of our models. The best ERP software providers need to know the value of being transparent regarding the accuracy and limitations of models. By creating trustworthy solutions that empower users to make informed decisions confidently, AI can truly make a significant difference to ERP processes.