How bots are relieving a global bank’s reporting woes . . .

Our client, a global diversified bank with operations in every major financial center in the world, wanted to make its legal entity reporting process more efficient. As it prepared to meet the demands of a growing list of reports, the institution knew it wasn't allowing employees to work efficiently.

Staff struggled with too many manual, repetitive tasks. Skilled professionals were handling transactional activities and high volumes of exceptions, leaving little time for analysis. Although the company wanted to automate many of these processes using its RPA platform, it needed custom-programmed bots to handle each task. But first it had to prioritize, optimize, and prepare the underlying processes for automation.

The firm's key challenges included:

Disparate data sources

Staff had to manually extract data from multiple sources using different formats. Then they had to transform and manipulate the data using ineffective tools, such as spreadsheets, and upload content into reporting systems.

Inadequate data quality

The team had the arduous task of individually resolving data quality issues and designing workarounds. It was the only way to meet the complex requirements of multiple legal entities and different reporting systems.

Obstacles to standardization

The bank wanted greater consistency, but requirements and processes varied across products and entities, making it difficult to apply consistent approaches and implement robust controls.

For digital transformation to succeed, employees need to be supportive. Genpact worked in partnership with employees across departments including IT, the central RPA team, risk and controls, and finance to identify internal change champions. Their role was to educate other employees about the benefits of automation. Following a bot demonstration to show how the new approach would improve their work, employees were happy to put aside their fears and false perceptions.

With employees backing the project, Genpact took a three-step approach: assess, design, and build.

Assess

We applied Genpact's Intelligent Automation Index to pinpoint where automation had the most potential in the LER function. The assessment considered many aspects of existing processes, such as the levels of standardization, criticality, and dependence on IT systems. While highlighting the potential for RPA, it also examined the other building blocks of intelligent automation, including artificial intelligence (AI).

Design

Our subject matter experts collaborated with stakeholders across different entities and functions to standardize and simplify an array of existing reporting systems. We identified the processes with the most potential for simplification and automation, including opportunities for straight-through processing, while also prioritizing those with the best ROI. Another goal was to identify clusters of activities in which robotic code could be replicated and reused.

We identified five key activity clusters that were right for automation:

• Extraction of data from feeder systems
• Reconciliation of data across reporting systems
• Data enrichment using standard assumptions
• Exception handling
• Uploading of enriched data to the reporting tool, and real-time dashboards for financial control

By pinpointing the tasks within each cluster in the RPA platform, we were able to quickly replicate them across different reports and accelerate bot rollout. Also, by leveraging the computational power of the bots, we were able to significantly augment the bank's overall operational controls.

Build

We implemented the process re-engineering strategy that we identified in the design phase by creating RPA objects that could be replicated across various reports, products, and organizational entities. As we deployed new bots, each one was rigorously tested, first by our technical experts and then though a systematic program of user acceptance testing. When the automated processes went live, we activated our hypercare maintenance framework to handle ongoing change requests and stay compliant with service-level agreements.

As robotics automates financial report production, bots work in a hierarchy of supervisor and worker roles. Supervisor bots allocate and manage work based on bot availability, how critical the activity is, and what other tasks depend on it. Each worker bot has a profile in the enterprise resource planning system that the supervisor uses when allocating work.

This holistic, end-to-end approach establishes the most effective scope for RPA implementation. It also allocates systems to clusters where bots can be replicated while meeting the individual needs of each process and report. This approach guided the creation of replicable custom-coded objects for the RPA platform, which reduced the development effort considerably. With 20 usable codes, we reduced the effort spent on development and testing by around 15%.

This project demonstrates that RPA does not remove the human element from an organization. Rather, bots complement and elevate human expertise throughout the process in smarter, more comprehensive ways.

The 61 bots Genpact programmed and delivered remain hard at work for the bank, all day, every day, overseeing rule-based repetitive steps and eliminating tedious tasks, which frees employees to focus on higher-value work.

The tactical benefits are:

• Automation of up to 60% of manual tasks
• Process cycle times up to 75% faster
• Over 90% fewer exceptions caused by data quality issues
• Full time equivalent capacity increase of 30% due to bot-boosted productivity
• Enhanced process transparency and audits
• Clear dashboards to highlight changes made and flag the exceptions the team must deal with

The impact of RPA deployment goes beyond the reporting function, as appetite for digital transformation is growing across various processes. Genpact continues to work with the finance leadership team to deliver further innovation.

The next steps will see the financial services firm's introduction of technologies such as AI and machine learning to enable intelligent automation.