RPA for Call Monitoring

Problem Statement:

Company provides services to their clients for call monitoring. The customer support center team of these clients gets daily calls from existing or new customers Multiple calls are received in a day and every client wants to have a summary about all calls, their type, reason of call and next action required for each call at the end of the day. It is very difficult to get this data which is very crucial for company for maintaining service and quality. Here, Company provides a service to monitor daily calls and prepare the desired summary report for each client. For this, they have a team who listens to calls and extract the desired data for each call and feed in a web application. This task is a repetitive and time consuming in nature. To address this and improve the operational efficiency, Company decided to explore the benefits of Robotic Process Automation (RPA).

As stated, Company was grappling with challenges in its call monitoring services. The existing manual processes for monitoring and evaluating customer calls were labor-intensive, time-consuming, and prone to human errors. With the increasing volume of calls and the need for real-time analysis, company faced a critical need to enhance the efficiency, scalability, and accuracy of its call monitoring operations.

Solution Overview:

Robotic Process Automation (RPA) has gained considerable momentum within organizational practice in recent times. Organizations competing in multiple industries have experimented with or implemented RPA solutions to address time-consuming, repetitive, and error-prone manual processes in order to increase productivity, minimize costs, and enhance customer service. Many commercial vendors and opensource communities provide RPA tools such as Blue Prism, Automation Anywhere, UiPath, TagUI, and RPA Express. RPA is capable of handling both external activities (e.g., customer-facing, supplier facing) and internal activities (e.g., back-office functions, cross-functional communications, intra-departmental tasks) within organizations. Since business processes are structured sets of work activity that are unique to organizations and the systems used to enable such processes are custom-designed or customized for contexts, there is considerable variability in business processes across organizations.

To address the challenges faced in call monitoring, GSS opted for the implementation of Robotic Process Automation (RPA). The key objectives were to:

  • Automate Call Recording Retrieval: RPA bots were deployed to automatically extract recorded calls for each client from the system, eliminating the need for manual intervention. RPA logins on system using their employee login credentials.
  • Speech-to-Text Conversion & Data Extraction: Bots are integrated with AI model and were employed to convert speech into text transcript, facilitating efficient analysis of call content and extract the required data.
  • Quality Metrics Assessment: RPA algorithms were implemented to assess calls based on predefined quality metrics, identifying areas for improvement.
  • Data Entry: The bots are created to perform data entry into the system according to predefined quality standards. If the extracted data meets the specified criteria, the RPA algorithm reintegrates the data into the system.
  • Real-time Alerts: The RPA solution enabled real-time alerts for immediate attention to critical issues during whole process cycle, enhancing responsiveness.
  • Scalable Architecture: The RPA system was designed with scalability in mind, ensuring it could handle increasing call volumes seamlessly.

Tech Stack Leveraged:

RPA is a software solution that uses “bots”—i.e., software configured to handle specific work previously done by humans—to automate business processes. It strives to mimic how humans may complete activities in digital processes using the same systems and user interfaces. Bots are able to extract data from multiple sources, process such data using well-defined rules, and generate desired output that may be directed to other systems. By performing routine, rules-based service processes that lead to deterministic outcomes, bots can complement human employees such that human-robot teams may eventually yield significant benefits such as reductions in human error and staffing costs. RPA solutions offer considerable advantages since they are essentially middleware systems that are configured or developed without altering the existing enterprise platforms and applications within the implementing organization.

TagUI is an open-source, cross-platform, command-line RPA tool that allows you to automate your desktop web, mouse, and keyboard actions easily. The automation software uses “human language” like Command line syntax to build your workflows. It helps you rapidly automate your repetitive or time-critical tasks – use cases include process automation, data acquisition and testing of web apps. The goal of user interface (UI) automation is to reproduce cognitive interactions that you have with websites or on your desktop. This is so that your computer can do it for you based on your schedule or conditions. Here are some key features:

  • Human readable code – can be written in 20 languages
  • Screen-based web automation possible using computer vision
  • Can integrate with AI/ML using API calls, Python
  • For specific requirements, JavaScript, XPath, HTML is required
  • Live mode for development and debugging. See outputs as you type
  • Simple installation: Just unpack and run
  • Automate Chrome in visible/invisible mode
  • Chrome extension for recording web actions
  • Python & R integration for big data / AI / ML

Technology stack for this deployment includes:

  • RPA Platform (TagUI): TagUI is a popular RPA platform that provide user-friendly interfaces for designing automation workflows, integrating AI capabilities, and managing bots efficiently.
  • Speech to Text Conversion: Whisper model provides accurate and scalable speech-to-text conversion, providing the text data for further analysis.
  • Quality Metrics Assessment: Custom algorithms tailored to the specific quality metrics like confidence score calculation for each call, implemented using a programming language Python.
  • Scalable Architecture: Containerization using Docker provides a scalable and flexible architecture for deploying and managing RPA bots.
  • Data Storage: Storing extracted data in a MySQL database for easy retrieval and integration back into the system.
  • Development and Scripting: The implementation of the necessary solution involves the use of various languages, including Python, Node.js, and the Laravel framework. Python is employed for training AI models, Node.js is utilized for crafting custom algorithms and RPA scripting, while Laravel is chosen for constructing the user interface to monitor RPA processes.
  • Real Time Alerts / Notifications: Using email alerts, system sends real-time alerts to supervisors or relevant stakeholders. It includes performance reports, error logs and failure incidences.

Benefits delivered:

  • Operational Efficiency: RPA significantly reduced manual effort in call monitoring, leading to increased operational efficiency and quicker response times.
  • Scalability: The automated system easily accommodated growing call volumes, ensuring that quality monitoring services could scale with the business.
  • Real-time Insights: RPA facilitated real-time analysis, empowering supervisors to promptly address emerging issues during operations, improving overall service quality.
  • Cost Savings: Automation of repetitive tasks resulted in cost savings, allowing ABC company to allocate resources more strategically and efficiently.
  • Improved Quality Assurance: The RPA-driven call monitoring system enhanced the quality of customer interactions, leading to higher customer satisfaction.
  • Agile Response: Real-time alerts and analytics empowered supervisors to respond swiftly to emerging issues during live calls, improving customer experience.