Intelligent Process Automation (IPA) is a powerful approach to rapidly automate business processes and achieve key business goals like cost reduction, operational efficiency, and error reduction. It combines technologies like robotic process automation (RPA) with artificial intelligence/machine learning (AI/ML) technologies like optical character recognition (OCR) and natural language processing (NLP) to automate as many steps as possible in a business process workflow.
Popular IPA platforms like UIPath, Blue Prism, and Automation Anywhere enable rapid development and deployment of automated processes. Multiple organizations have piloted IPA programs and achieved key business objectives in less than a year.
Process mining is a data-driven approach to capturing information about a process and the variations in its workflows. The traditional method for capturing process automation information is usually done through process walkthrough sessions and interviews with operation teams.
Process mining works by extracting data from logs, database records, and other application data sources and automatically building maps representing all process variations. It can complement manually created process maps to provide a more comprehensive end-to-end view of a process. A process mining platform can also analyze process data to automatically calculate key KPIs and pinpoint improvement opportunities by elimination, optimization, or automation. It can also continuously monitor live processes and track progress in transformation initiatives.
Process mining can be applied at multiple stages of an IPA program. The two technologies enhance value across the automation lifecycle significantly.
Pre-automation
Process mining can help build and improve the quality and efficiency of the RPA requirement capture and solution design process. The pre-automation stages include:
- Identifying automation candidates:
Process mining platforms can capture data points on process FTE count, manual effort, current automation levels, volume, frequency, and error rates. The platform automates the discovery of processes and sub-processes with high automation potential. Also, process variations and steps with long average cycle times and other inefficiencies are analyzed in detail to design the optimization or automation approach. Additionally, the business case can also be formulated. Automation candidates can thus be selected and validated using objective criteria.
- Requirement capture:
Process mining can augment interview-based requirement capture by automatically generating process maps from system records and event logs. These can be used to validate the information captured in requirement sessions and add additional workflow variations. Leading process mining platforms like Celonis have task mining capabilities to capture detailed operator click levels and ensure that the requirement capture exercise is broad and exhaustive.
- Solution design:
Leading process mining platforms have process modeling capabilities to help design the future state of the automated process. The behavior of the future process can be simulated under various input conditions to predict the future benefits of automation. Processing mining-driven inputs can enhance the solution design exercise and help identify inefficient sub-processes that can be eliminated or optimized. Additionally, the solution can be enhanced, and AI/analytics models can be built on the platform to trigger a bot under certain conditions.
- Test design:
Information on multiple variations of the process captured by the process mining platform is used to design test plans and cases. This will ensure comprehensive test coverage and a robust quality assurance cycle when developing the automation solution.
Post-automation
Process mining can help track the success criteria of automation progress and be used to improve the performance of support operations after the solution is live. The post-automation stages include:
- Measure transformation goals:
The process mining engine can continuously track key performance indicators (KPIs) in an automation program. Efficiency improvement, cost reduction, and error reduction can be constantly tracked. Also, leading process mining platforms have benchmarking features and can use this to compare the performance of automated processes against similar processes in the organization.
- Capture additional opportunities:
There can be additional opportunities for optimization and automation of an automated process. A process mining tool can continuously analyze the bot execution workflow and discover new opportunities.
- Bot monitoring and support:
Process mining platforms can connect to live bots and their infrastructure to monitor and track bot uptime, health, and failure rate. These platforms' predictive analytics and AI capabilities can even be leveraged to predict bot failures and take action to prevent incidents.
Process mining is a powerful enabler of success in intelligent process programs. It can ensure that the right processes are automated and mitigate challenges commonly faced during an automation project's planning and implementation stages. It can also drive the effectiveness of post-automation governance and benefits realization.