Back-office innovations improving mission performance

Trend in action

Back-office innovations and mission tasks are converging

As both mission and back-office tasks become increasingly data-focused, improvements to one area can benefit the other. Many government agencies across sectors are using data to innovate back-office functions to allocate resources, identify risks, and improve service delivery.

Allocate resources: Components within the US National Institute of Health (NIH), the biggest funder of biomedical research in the world, deployed artificial intelligence (AI) to assist in assigning grant applications to appropriate review groups.¹ Prior to this, staff would manually read applications to sort them by scientific discipline—a process that took several weeks and delayed the referral process. The AI tool scans an application’s text, title, and abstract and assigns it to an appropriate group with 92% accuracy.² This innovation has reduced the time taken to process grant application from 2–3 weeks to less than a day, ultimately accelerating NIH’s core mission of making scientific discoveries.³

Identify risks: Transport Canada, the department responsible for regulating transportation, receives nearly one million preload air cargo records a year. It began requesting these records after authorities (acting on a tip-off) intercepted two bombs hidden inside LaserJet printers shipped in 2010.⁴ Transport Canada hoped to identify other potentially dangerous cargo early. Reading manifests by hand simply wouldn’t work.⁵

An AI tool came to the rescue. Using previous air cargo records and the results of manual risk assessments, Transport Canada trained an AI model to rank cargos by likely level of risk. In parallel, they trained it with natural language processing to sort cargo into meaningful categories. Combining these capabilities, the tool collects detailed cargo data, detects anomalies, and identifies unusual patterns for staffers to evaluate.⁶ The back-office task of reading manifests not only speeds processing, but also helps protect Canada’s air, rail, and other freight.

Improve service delivery: Trelleborg, Sweden, has used AI to automate various social assistance decisions. The automated decision-making system has reduced the time to process applications for home care and sickness and unemployment benefits from 10 days to less than 24 hours. Caseworkers review applications the system rejects to ensure the AI has not made an incorrect decision.⁷

Reimagine operations: Cascais, Portugal, reimagined its operating model through digital twins by integrating data from multiple verticals, ranging from health and transportation to energy and public infrastructure. Implementation of their smart waste-management system combined with real-time traffic data allowed the city to optimize routes, identify best times to collect waste, and reduce operating costs by 40%.⁸

Breaking down barriers between back-office and mission innovation changes how value is created

Governments are recalibrating. To make the most of new technologies, back offices should streamline workflows. A Deloitte AI survey revealed that organizations that have significantly changed workflows are 36% more likely to achieve desired outcomes from their AI projects.⁹

The United Kingdom’s “Tell Us Once” program allows residents to notify tax authorities, the passport office, local governments, and benefits programs with a single click, instead of notifying each agency separately. The program rests on a new application programming interface (API) that allows for the sharing of information between agencies while still preserving citizen privacy. But this new technology needed to be complemented with new business processes within each agency to take advantage of it. The benefits of such changes may be far-reaching. Not only can it reduce citizens’ efforts, but the program also saves the UK government £20 million annually.¹⁰

Saving money is not the only way to quantify the value of an innovation. Breaking down the barriers between back office and mission also mean that improvement to mission can be an important way to quantify the value of an innovation. For example, moving to an intelligent digital process helped the US Department of Education improve the application process for student aid. The redesigned back end allowed students to pull tax information directly from the Internal Revenue Service systems and include it in the application. This redesign was designed to simplify the process, increase accuracy, and reduce improper payments. Furthermore, the application saves students’ personal information, so they don’t have to reenter it when they apply the next year.¹¹

But new innovations creating value in new ways often won’t work with old business processes. Rather, creating new processes can better fit innovations to the value they create, increasing time savings and reducing backlogs. In 2016, Azerbaijan automated a system to clear 165,000 uncontested court cases. Just as some automatic processes must be integrated with current back-end systems, this one had to be integrated with the legal system. New legislation was needed to allow for electronic filing of court documents and centralized management of those cases. With that legislation in place, the pilot could proceed, and the average time to clear a case dropped from three days to one.¹²

These shifts change how work is done and who is doing it

Governments can tap into external ecosystems to reimagine their back office. Many of the capabilities and infrastructure to support data-driven decisions may reside outside of government.

During the pandemic, demands on government services increased and changed how citizens requested those services. People applied online for services they used to wait in line for. Many government agencies turned to the private sector to upgrade their technology systems to meet demand. In our 2021 global digital survey of government officials, 85% of respondents said that collaborating with external partners positively impacted their agencies. More than 80% also indicated the same for the use of contractors.¹³

VITAL, a government shared-services agency in Singapore, is launching a cloud-based platform that would make robotic process automation (RPA) a norm for finance, HR, and procurement services. Agencies are expected to be able to use the platform to generate reports from extracted data, send automated notifications, and reconcile data and information through software bots.¹⁴ For recruiting, VITAL partnered with an external AI provider to enhance the existing process. The AI solution can help agencies incorporate inclusive hiring processes by masking personal information, automate screening and matching, and verify candidate documents. It ultimately is designed to help multiple agencies reduce the time and cost to hire talent.¹⁵

Engaging the right partners can also reduce the risks of modernization. The US Food and Drug Administration (FDA) used a managed-services approach to automate its highly complex and manual invoice processing. The FDA used RPA bots to input data and perform calculations on the invoices, some of which were 50 pages long with more than 500 line items. The bots sped up the process and reduced human error. The automation is designed to help the agency meet governmentwide objectives of modernization, improve transparency, and accountability to Congress and taxpayers.¹⁶ The FDA’s choice to use managed services meant that it could quickly deploy the bots without having to train or hire a dedicated workforce. The managed-service contract also covers continuous maintenance of software and retraining of bots to respond to future conditions.¹⁷