Robots: an added value in Fund Administration


As part of the Societe Generale group’s digital transformation, the Securities Services business started using Robotic Process Automation technology (RPA) in 2016 to improve its production processes and make operational gains. A report on the feedback and lessons learned.

In the second quarter of 2016, and with the help of consulting firms, SGSS launched its first Robotic Process Automation (RPA) developments in Europe (France, Luxembourg and Italy). When the project began, SGSS adopted BluePrism software, but this was quickly replaced with the rival solution, Automation Anywhere, which is thought to be more suitable for users with limited IT skills.

The first processes targeted by these developments were the entire portfolio valuation chain, under the responsibility of the fund administrator, and some processes for managing orders handled by the registrar. At the same time, SGSS launched a project for its fund administration work in India, where there is a high percentage of manual labour combined with high volumes of operations.

The lessons that SGSS has learned from these developments

Feedback and lessons learned relate to two aspects; one operational, the other technical.


Since 2017, SGSS has introduced more than 30 robots around the world, the majority of them in India where processes are simpler and involve higher trade volumes. These developments were applied to various processes such as asset and liability order management, flow integration, NAV calculation control, etc.

At the beginning of 2017, on the basis of this experience, SGSS was able to draw its first conclusions. RPA seems to be a technology suited to processes with the following characteristics:

  • high volume of manual processing;
  • little business complexity;
  • limited processing rules.

Conversely, robots are definitely not suitable for automating the whole production chain. Indeed, the diversity of specificities within a processing chain does not yet allow a robot, as powerful as it is, to automate these rules. Laying down a series of rules for business lines is considered too expensive in testing and maintenance terms.


As with any project, the creation of a project management unit specialising in RPA was needed to be able to produce business requirements and operational specifications for future developments. The organisational aspect also had to be studied to establish an efficient operating model (application support, updating of new app versions downstream, etc.)

Despite what software publishers and consultants may say, the IT department’s involvement is essential for this type of project, mainly due to the infrastructure problems inherent to the use of these technologies. Such projects cannot be managed by the business line alone. Indeed, the use of virtual machines to install internationally available robots or apply the password security solutions used by robots are just some examples of the many challenges that must be overcome. It is also important to stress that RPA technology does not appear to be sufficiently mature for applications available remotely (through CITRIX), given the screen recognition problems that they can present.

Robots are definitely not suitable for automating the whole production chain.

Our experience suggests that use of RPA for SGSS’ registrar and fund administration businesses, which traditional developments have largely automated already, is only suited to high-volume manual processes. Whenever clients’ specific needs, complex management rules or low volumes are requested, the return on investment in these developments is low. Moreover, RPA cannot meet current automation without being integrated with other systems.

Robotic Process Automation: a necessary first step towards better-performing technology?

The results obtained so far show the need to go further to truly realise the promises of savings already potentially obtained at this stage. SGSS is now exploring new production chain opportunities combining RPA with technologies such as optical character recognition and machine learning to enhance processes that are yet to be automated.The introduction of these robots is therefore a first pivotal step for SGSS in its plan of rolling out new technology to digitalise some of its operations.

Deputy Head of Transformation Societe Generale Securities Services
Did you like this article ?