There is an ideal of the future-facing professional accountant. A truly cutting-edge professional with a broad skill set that includes technical prowess, technological savvy, excellent communication skills, untouchable ethics, commercial acumen, a focus on sustainability and governance advocacy.
For many, the future has arrived and all these qualities are expected of finance professionals in the current job market. This is evident in companies seeking increasingly advanced digital and data skillsets that include machine learning, AI, and programming languages such as Python.
A quick job search on ACCA Careers using the keywords ‘machine learning’ uncovers roles such as digital audit advanced analytics senior associate at PwC; assurance – forensic & integrity services manager at EY; business analytics analyst at DBS Bank; senior data scientist at KPMG.
These roles variously ask for people with light experience or knowledge of data analytics and visualisation software and machine learning, all the way to proficiency in programming languages used to query data such as Python and SQL, or create visualisations using PowerBI or Tableau.
Accountant or data scientist?
While we’re perhaps not at the stage yet where finance professionals are expected to be data scientists fluent in computer programming languages, accountants are increasingly working alongside such professionals to deliver data-driven management information, and to tell better stories with a business’s data.
And it’s entirely feasible to expect that in years to come finance professionals with programming ability and deep data science skills will either be the norm or at least highly desirable to business.
This is one of the drivers behind ACCA’s decision to launch a new online course run with EdX called Machine Learning with Python for finance professionals. The course provides a platform for an effective and productive understanding of machine learning and languages like Python, which will better enable those who sign up to interrogate data models and output and work effectively alongside data scientists as partners.
Unstoppable innovation requires continuous learning
For Tee Chong Yu FCCA, audit analytics senior manager at Deloitte Singapore, the stories that data can tell are endless. ‘I get a sense of satisfaction when I see the amazement on my colleagues’ or clients’ faces when I tell them a story they’ve known all along, but never had the data to back it up.’
For a role like Chong Yu’s in audit, which is a field of the profession seeing distinct technological disruption, an ability to code in SQL, Python, or R along with knowledge of visualisation tools are considered necessary.
However, he notes that having the attitude and aptitude to learn and re-learn is also incredibly important. ‘Learning because there is a multitude of tools being created on an ongoing basis, and re-learning because as we analyse a given set of data, the previous assumptions that we held true may be dispelled due to circumstances relevant to that data set.’
Machine learning and accountants
Fortunately for modern professionals, machine learning will automate processes and free people up to add the kind of rich value machines can’t and that heavily manual and repetitive tasks wouldn’t allow them the time to.
The point of machine learning is to write an algorithm that can be trained using test data to look for specific patterns. An example of this is Xero’s invoicing software in which the machine learns how a business is run and automatically completes invoices based on previous entries.
Other examples include machine learning automating bank reconciliations: technology learns from previous allocations and account choices and makes the right recommendations for new bank transactions.
Automation means auditors are starting to audit 100% of a companies’ financial transactions, whereas in the past due to the huge volumes this would require thousands of hours and overtime. Machine learning algorithms will process and review data, catch anomalies and produce lists of outliers for auditors to check. This means auditors will spend less time checking data and more time investigating and questioning patterns or anomalies.
Python in action
Alastair Barlow FCCA is the co-founder of flinder, a digital-first practice that employs a data operations team of data scientists that works in partnership with the firm’s accountants. ‘In our business, the accountants don’t need to understand the black box, but they do need to understand the inputs, basic process, and outputs as they own this for our clients.’
flinder uses Python to manipulate and visually report on data from large and multiple data sets. ‘It’s an extremely effective way to identify trends and meaning from the data that can then frame actions. In addition, we use it to automate manual activities which fall outside cloud applications,’ said Barlow.
‘Basically, it can increase efficiency, increase control and significantly increase insight.’
flinder uses Python in four main areas:
- Integrations: reduced human intervention in processes between applications
- Data manipulation: significantly reduces the time to process/review data sets and provide exception reporting
- Data visualisation: pulls huge volumes of data together to tell a concise story across the population and specific sub sets of the population
- Automation: developing scripts to automate specific tasks
As far as Python forming a part of an accountant skillset, Barlow thinks it would be an effective addition to an accountant’s core skillset. ‘I think over time we will definitely have accountants that use Python, but this will form a more specialist niche skill in the future, much how a specific type of tax is now, rather than all accountants or a generalist working with it.’
Link: Machine learning with Python for finance professionals
Author: Neil Johnson
