Problem Statement

Approximately 60% of engineering effort was observed to be spent on product maintenance in a quarter, leaving very little availability for investment in long-term initiatives.

Context

• In the Order to Cash lifecycle of a B2B transaction, Robotic Process Automation (RPA) bots automate data aggregation from end-customers. Click Here to learn more about the Order to Cash process
• RPA bots are designed to aggregate data from two main sources:
  • Websites
  • Emails
• The process of RPA bots aggregating data from websites involves four steps: Login → Bot navigation on the web portal → Data Aggregation → Data indexing in cloud
• In the current state, the maintenance and modification of RPA (Robotic Process Automation) bots are highly code-driven, leading to:
  • Increased time for debugging and implementing changes.
  • High engineering effort devoted to fixing bugs and performing minor updates.
  • Delayed deployment of updates, affecting operational efficiency.

Challenge

<aside> ⚠️

1. Approximately 35% of websites where RPA bots were deployed experienced frequent navigation changes, causing temporary disruptions in automation workflows.
2. Slow Turnaround Time: Complex debugging and long deployment cycles delay updates and reduce the responsiveness of the RPA solution.
3. Dependency on Skilled Developers: The code-driven framework requires skilled developers with an awareness of fixing issues fast.
4. High Risk of Errors and Downtime: Manual code changes, if not thoroughly reviewed or tested due to backlog overload, increase the risk of introducing bugs, which can lead to downtime, data inaccuracies, and operational disruptions.

These challenges pointed to a long-term scaling issue, with an increasing number of bots to be added and maintained with a growing demand for engineering resources

</aside>

→ Based on the above analysis it was concluded that the current RPA framework needs a revamp in order to scale the product without growing investment in engineering effort.

Impacted Stakeholders

<aside> 🎯

• Deduction Analysts facing frequent interruptions and extended product downtime, leading to poor user experience and lower Net Promoter Scores (NPS).
• Account Receivables Managers encountering inaccurate financial data compilation due to missing information caused by automation breakdowns.
• Developers continually tasked with repairing RPA bots, resulting in decreased morale and reduced efficiency to resolve issues. </aside>

Business Objectives

• Goal 1: Reduce the quarterly engineering effort spend spent on bot build by at least 50%,
• Goal 2: Reduce the quarterly engineering effort spend spent on bot maintenance by at least 70%

Solution Approach

Solution Overview: Configuration-Based RPA Framework

The proposed solution is a configuration-based RPA framework, which allows engineers and, where possible, non-technical users to manage, update, and deploy bots through a user-friendly configuration interface instead of modifying the underlying code.

This approach separates process logic from code, reducing engineering effort, accelerating turnaround time for updates, and enabling scalable RPA operations.

Key Features for the MVP version of the framework