‘Treat CODA files and do process optimization’ was the initial request TriFinance received from one of the leading financial institutions in the field of instalment loans and revolving credits. At the end of the project, our consultant Marco had been in the lead of a big RPA project involving those CODA files and realized other process optimizations in the finance domain.
Master the basics to understand the needs
A CODA file (Coded Statement of Account) is a payment in the form of a coded bank statement that can’t be attributed directly to a corresponding credit contract due to the lack of a structured reference. These payments need to be allocated manually on a daily basis, which is a repetitive and time-consuming task. On average, it’s possible to treat 100 codas an hour. Imagine the pressure on a small accounting team when experiencing more than 2K of coda files with the daily deadline of finishing this treatment before 4 pm due to closure reasons.
The project's primary goal was to provide operational support in this matter on a daily basis. It provided me with critical insights and ideas, making it possible to participate in the RPA project that was still in an early stage at the start of the mission. Without those basics, bringing this challenge to a good end would have been tough.
Maintaining the balance
The goal seemed very easy: ‘create a bot which allocates CODA files automatically’. Reality demonstrated it was all about maintaining the balance between internal and external parameters. The script needed to be unambiguous and efficient. AS400 (i.e. the program used to treat the payments) had its own particularities that needed to be taken into account.
The pace of treatment was also an important factor, as more and complex rules implied a lower average treatment rate. The impact on other accounting processes (e.g. ‘lettrage’) needed to be examined as well to avoid disruptions.
With profound analysis, a lot of testing and clear communication, we managed to achieve the ideal ‘mix’ in which we have created a bot that is now fully integrated to reduce the daily workload significantly. It’s not only the average allocation rate of +40% which makes this story a success but also the ability of the bot to increase its treated numbers on the days when the amount of CODA files is at its highest.
With profound analysis, a lot of testing and clear communication, we managed to achieve the ideal ‘mix’ to create a bot able to significantly reduce the daily workload.
Change management and automation go hand in hand
As this was one of the first robotizations in the firm, there were a lot of eyes on this project, not least from the people who were directly impacted in their day-to-day activities. Scepticism and uncertainty were among the negative emotions we encountered during the development stage.
We tackled this and other opposition by directly involving the people in the project, hereby stressing the importance of their feedback. We emphasized the gains and the new opportunities that will be created. There is now less repetitive work to be done and more time to be spent on other more challenging and interesting tasks or projects. Only when a project like this is supported by the people, can we talk about success.
RPA - Project Approach
The principles of the Waterfall approach were followed to a great extent during this project. There was a lot of aftercare needed to point the bot in the final direction. Most of the issues arose from the program used to treat the coda files. As the testing part was started after the complete analysis and development stages, I wondered if it would have benefited the project if Agile principles were applied. If small teams of business analysts, developers and testers had worked closely together in sprints, maybe the current ideas which are bundled for a second version of the bot would already been implemented and the aftercare could have been reduced to a minimum. Perhaps a more agile way of working could have increased the efficiency in this project and hence increased the ability to scale if future RPA projects see the light in this institution.
Personal development
The least I can say is that I started this project from zero. With no experience in process management and scarce knowledge in the Finance domain, I managed to enlarge and enrich the initial scope of the mission by taking initiative and challenging myself in different ways. Thanks to strong mentorship, I was able to seize and create new opportunities during this mission.
I improved my communication skills in listening and discussing different ideas on the topic with several stakeholders to make informed decisions. The project showed me that it is essential to take a step back to get a global view of all the realizations and ensure you’re still on the right track before proceeding to the next stage. Because of the constantly changing Covid-situation, I learned to adapt quickly and be creative when necessary.
This experience has triggered me to know more about RPA in Finance and other domains. I want to follow this path further, leading to efficiency in future projects.
This project is a concrete example of what TriFinance stands for: a hands-on and pragmatic approach starting from a good knowledge of the underlying business.
Conclusion
There is a lot of automation potential in the Finance domain, not only in the field of RPA but also in, e.g. OCR technology (Optical Character Recognition). As stated above, it’s crucial that you know your subject thoroughly, as obvious as it may sound. The aspect of managing change needs to be well embedded in an automation project to succeed fully at the end of the ride.
This project is a concrete example of what TriFinance stands for: a hands-on and pragmatic approach starting from a good knowledge of the underlying business to implement the proper process optimization & automation solutions.
More information on RPA at TriFinance
Our holistic, process-first and people-driven approach to RPA focuses on automating routine processes, lowering costs and freeing up time for your employees to focus on value-added tasks.
