The world around us is changing fast, and as an investor and a company, we want to be able to respond to this well. We have a clear goal to be best in class. The realisation of this goal is inseparably connected with the use of data and AI. However, we were trailing in this field. There was insufficient fundamental attention to data within PGGM. As a result, there was a lack of insight into the possibilities that use of data and AI entail. PGGM also did not have a modern infrastructure for working with data and AI.
We want to change this situation. In the past two years, we have realised many changes in the field of data and AI use. We formulated a data policy and, even more importantly, took a number of concrete important steps to realise that policy and to take advantage of the benefits of data and AI use in business operations.
Why AI is so important
We see an important role for AI in our working processes. We therefore devote growing attention to this development. AI enables shorter throughput times, faster gathering of information and better processing of information. A number of concrete examples of how AI can contribute towards daily processes in the near future are presented below:
AI can provide support through a better insight into transaction costs, both predictively and on the basis of historic pattern recognition.
Identification of the impact of investments on SDGs.
Support for and automation of infrastructure processes (in particular for Listed Real Estate).
Within the Risk department, AI can provide support by making market trends and deviations from the standard (in the case of public investments) transparent faster and by providing understanding of and insight into information for private investments.
In the back office, AI can be relevant by automatically providing texts and explanations for reports, possibly even in dialogue with the reader.
Making investment proposals involves dealing with an enormous amount of data. AI can help here by summarising information and making it more transparent.
Even for a mandate with automated rebalancing, the data management must be in order: AI can provide an insight into data quality.
PFZW invests in a range of different asset categories. For private markets, speed in decision-making on deals is extremely important. AI can help here by making the enormous volume of information transparent.
AI contributes towards our strategic goal of making the organisation more flexible and efficient and helps us in our drive for best-in-class asset management. We will develop this further in 2020. Some examples of processes in the field of AI and robotization that are now in use at PGGM Vermogensbeheer are presented below.
Concrete approach from Datalab
PGGM is increasingly developing into a data-driven organisation. Data is an integral part of our strategy and our future service provision. Within PGGM, we continually consider how we can use pension- and asset management data safely to improve our service provision or to develop new services. The PGGM Datalab plays an important role in facilitating this process still better. The Datalab approach is distinguished by the deployment of a multi-disciplinary team and a use case-driven approach.
The deployment of a multi-disciplinary team means that not only are data scientists needed for the realisation of a data-driven approach, but also the deployment of a whole series of other disciplines and roles, such as data-engineering, cloud service engineering, security, legal services, compliance, procurement and various trainers. Datalab has brought together all of these disciplines in a single team and deployed it for the realisation of use cases. That was the condition for realising results quickly and for exposing gaps in the infrastructure. Thanks to this approach, Datalab has laid a new foundation and implemented infrastructure, including stakeholder management and permission from our client PFZW to move into the cloud. Through ‘use case-driven’ operations, the Datalab has addressed and realised dozens of cases, together with colleagues from seven varied teams. In this way, together with the Datalab, we have gained experience with data and AI and have shown the added value through use of data and AI. At the same time, Datalab has organised training courses for employees, so that they gain knowledge and experience relating to the use and the value of new data technologies.
Robotic Process Automation
Within PGGM Vermogensbeheer, we have been actively working on Robotic Process Automation (RPA) since the end of 2018. At present, 24 robots are in production, spread over seven different departments.
RPA is used to automate our business processes and so to improve service provision for our clients. This takes place with the aid of software robots. The robot communicates with the user interface of the relevant application. As a result, a robot can replicate human actions.
A large proportion of these robots are used in the processing of the cost administration. One example of this is a robot that helps in checking invoices of External Managers. Invoice amounts are compared with contractual calculations and if they fall within the margins, the invoice is paid.
Another robot that has realised many savings is the Holding Reconciliation robot. This searches for differences in holding positions by checking the transaction lists. At the same time, it shows the classifications of the transactions.
The objective for the coming year is to automate all repetitive and time-consuming activities within the asset management unit, so that employees can focus on the analytical work.
We started work on the Datascience Bootcamp in 2019. In this programme, we teach our employees programming in the Python programming language. This gives them leads with which they can build AI models. They are then able to identify and apply the opportunities that AI entails (they do this in collaboration with the Innovation department and the PGGM Datalab). The programme is divided into different modules. Several dozen employees took part in each module in 2019. In 2020, two modules will be included in the offer of PGGM Academy, which will make them accessible to all PGGM employees.