DUBAI - The Ministry of Finance has just completed phase two of embedding robotic process automation (RPA) into many of its internal processes.

The Ministry now uses bots, software applications that carry out automated tasks for 1.8 million transactions with greater than 98% accuracy, saving 39,000 hours of human labour.

Bots are also increasingly used in the public sector as government entities try to improve efficiency and customer services.

Before 2020, the Ministry identified functions that could be improved with automation. These included accounts payable, payroll and pensions, financial accounting, IT processes, web services, and hardware monitoring.

By 2022, it had more than 50 bots working unattended, delivering great results, and it expects to save an additional 12,000 working hours in 2023. To date, the Ministry of Finance has automated 63 processes and subprocesses, reflecting a 95% reduction in errors and a 65% reduction in average handling time.

RPA is one aspect of modernising government affairs, creating a culture of excellence. In turn, the culture creates a conducive environment for innovation, which paves the road for the transformation of the UAE as laid out in the ‘We the UAE 2031’ vision and the UAE Centennial Plan 2071.

The process automation initiative falls under the Ministry of Finance’s Strategic Plan 2023-2026, which is a roadmap to accelerate government performance through financial empowerment, sustainability, innovation, financial leadership, and sustainable development. It is underpinned by a series of pillars designed to improve the function of government by promoting innovative practices – such as the adoption of RPA.

On its website, the Ministry of Finance offers a Digital Procurement Platform, which enables it to accelerate processes from 60 days to six minutes, creating greater competition by allowing more small companies to compete for government contracts.

Phase three of embedding RPA is expected to raise the total of automated processes and subprocesses to more than 100, and a further reduction in average handling time by 10% with error reduction across all bot-handled processes by 98%.