Digital audit — providing more added value with new techniques and through new skills
In a data-driven society, the private audit sector needs to follow suit, because their clients increasingly ask for innovative added-value services made possible at low cost by new technologies. What challenges does a big private audit firm such as Deloitte see, and what features does it use and already offer to its clients? Ferry Meuldijk is Audit Innovation Manager in the Professional Practice Department — Innovations & Analytics at Deloitte in the Netherlands. He highlights some of the key techniques Deloitte Netherlands is using, but he also stresses that ultimately, professional judgement remains the pivotal element in an auditor‘s work.
By Ferry Meuldijk, Deloitte Accountants Nederland
Digitalisation is changing professions of all kinds, not just audit
The Fourth Industrial Revolution will create many new jobs within our profession in the coming years. These new jobs do require rather specific profiles. On the other hand, we also face the disappearance of an enormous amount of jobs as a result of new technologies. This threat exists mainly when it comes to accounting activities with a highly routine character, such as copying and reconciling data.
This trend is not only visible in our profession. The private sector is generally expected to be hit hard as it has higher rising staff costs than other sectors , such as the public sector or the non-profit sector. This makes the use of robots even more interesting, because they can easily take over processes in which many repetitive tasks are performed.
More and more employees are becoming aware of the fact that their profession is changing. In our sector, there is a huge demand for additional training such as Digital Accounting and we see an increasing amount of colleagues showing an interest in joining in our innovation and analytics projects during low season. We invite our colleagues to become frontrunners by developing and adopting innovation and analytics solutions, thus building up the skillset needed for the auditor of the future. This is because we believe that the real change will take place in the profession within engagement teams. They can make a real impact by adopting innovative tools to develop their profession.
An audit is a traditional product within a static, regulated framework. Methodology and control mechanisms in use under current laws and regulations are outdated and based on concepts from a vastly different time. For example, our employees and clients have data analytics applications available to them which can provide very accurate predictions. However, the use of such solutions is held back by our regulators who are not developing at the same speed. The results of some of the most innovative applications are therefore only used to support traditional audit activities.
Stakeholders and their expectations change and there is a huge demand for an innovative audit approach. Digitalisation is currently embedded in business operations and has a serious impact on the object of research. The use of smart applications also increases the dynamics. With the use of innovative applications, auditors can now make an immediate impact in comparison with the early years of most auditors’ careers which then consisted mainly of traditional ‘box-ticking.’
Automating the Data Driven Audit
By using innovation solutions to read, interpret and process big data, a new risk is created: the reliability of these systems. Especially when they are intended to replace the current control mechanisms in the future. There are quite a few conditions that need to be met before an auditor can use automated solutions and applications in the field of data analytics. The main condition is that the current control process needs to be optimised.
The first step in this optimisation is to standardise all procedures and processes. Within our organisation, we work worldwide with the same applications, working papers and specific workflows. This can also be seen in the private sector, where there is a continuous urge to standardise processes so that collaboration can be used more efficiently, financial data exchanged in compliance with the Payment Services Directive 2 (PSD2) and blockchain used, this maximising the transparency of all relevant information. This development will change the angle for third party trust providers. This exemplifies our view that all parties involved in the audit process remain equally important, but are different from before.
The second step in our optimisation is to centralise specific work and/or processes. By setting up audit delivery centres, specialised departments are being created with a focus on a particular part of the audit. In this way, we increase quality but there are also numerous advantages when it comes to efficiency. By experimenting centrally over recent years with new innovative solutions in automation and data analytics on a small scale, several components have been automated successfully.
As a final step, we automate parts of the audit process and transform how we do our work by using existing technologies in a scaled-up way. Robotic Process Automation (RPA) is hot. Almost everyone is doing it. But we have learned that making a robot is not the challenge. The real challenge is building robots in a scalable and reusable way.
At Deloitte in the Netherlands, we have invested heavily on a robust target-operating model and have now established a Centre of Excellence. We are running a cloud-based RPA infrastructure and have incorporated the robots in our standardised processes. The infrastructure is fully compliant with Dutch regulatory provisions and the highest risk and quality standards for responsible robotics.
Intelligent Process Automation
Clearly, we have been learning along the way. RPA was a first step. Intelligent Process Automation (IPA) is a combination of technologies to automate and integrate our audit process. Combining our robots with an emerging set of new technologies such as Natural Language Processing allows us to perform end-to-end process automation.
Natural Language Processing is a form of Artificial Intelligence, which allows us to extract key audit information from documents. This is done by defining the data points that we wish to extract, teaching the system where to find them and then training the system on multiple documents. The software analyses data in a way that a human could not, recognising patterns in data and learning from past decisions to make increasingly intelligent choices.
How does this work? We have integrated this cognitive automation into the RPA infrastructure. The robot will pick-up the audit evidence from our management system or client portal and transfer the data to one of our Natural Language Processing solutions. The software automatically captures, classifies and extracts key audit data from semi-structured documents like invoices and bank statements. The robot creates a working paper with an overview of the extracted data and matches it with the sample selection or general ledger.
RPA support in speed and efficiency by mimicking human actions to reduce labour-intensive tasks, such as copying or re-typing data. The AI solution supports by extracting the relevant audit evidence but the auditor is needed for professional judgement.
Artificial Intelligence in audit
We have also deployed a smart software application that innovates document interrogation and analysis by adding the power of hundreds of virtual eyes to the team. Leveraging Artificial Intelligence, this software solution uses advanced Machine Learning techniques and Natural Language Processing to quickly process, highlight and extract key information from electronic documents while learning what matters most throughout the process to make more precise extraction suggestions in the future.
The solution visualises (1) the similarity of contracts in a population, (2) the frequency of certain extraction fields in a document and compares contracts to a standard contract and (3) highlights the differences. A powerful search and analysis tool that enables our teams to spot issues and trends across documents. This solution again supports the auditors and manual labour-intensive tasks have been replaced by a machine. We are now able to examine entire populations. However, the conclusion cannot and should not be drawn by the machine. We need the innovative auditor to interpret the outcome of the analyses and take action.
It is not only in the Interim and Year-End phases that automation and Artificial Intelligence are used to support the auditors. During the planning phase, data analytics solutions are used to import processes and financial information in real-time. We have created an Artificial Intelligence solution that can be used during our risk assessment.
New technologies making collective knowledge available to the auditor
How to unleash the collective knowledge of all auditors, to help them to define a strategic audit plan? Imagine if we could leverage the strategic audit plans from thousands of other engagements and offer that in an efficient way to our auditors, supporting and challenging them when creating your audit plan. We have built algorithms based on the historical data from engagement files. Based on engagement characteristics the solution suggest risks, challenges risk classifications and risk strategies like a personal assistant would do. It asks ‘how does your engagement compare to other engagements?’ enabling more consistent and higher quality risk assessments. It improves efficiency by providing input instantaneously, supporting auditors in the decision-making process.
We have long heard that technology will impact audit, accounting and professional services. New technologies are here and they are not going away. They are not there to replace you but to support you. It is the synergy between accountants and delivery centres embracing new technologies. All are needed. None are replaceable. Embracing new technologies and innovations allows us to reimagine our audits, aiming to provide a smarter, better, and different type of audit.
This article was first published on the 1/2020 issue of the ECA Journal. The contents of the interviews and the articles are the sole responsibility of the interviewees and authors and do not necessarily reflect the opinion of the European Court of Auditors.
