Process analysis for digital transformation
Executive summary
Process analysis is best understood as a disciplined examination of how work actually flows across activities, roles, systems, data, controls, and outcomes. In traditional quality and operations settings, it has been used to reduce variation, waste, delay, and defects. In digital transformation settings, however, its role is broader. It becomes the mechanism by which organizations decide what to standardize, redesign, automate, instrument, and continuously govern. Recent literature positions business process management as a material enabler of digital innovation and transformation rather than merely a documentation exercise (IBM, n.d.; Putra & Mahendrawathi, 2024). [1]
The report’s principal finding is that no single methodology is sufficient for contemporary transformation programs. Workshop-based approaches such as SIPOC, value stream mapping, Lean, Six Sigma, and BPMN remain valuable because they provide shared language, explicit assumptions, and improvement structure. However, they should be complemented by data-driven methods such as process mining, object-centric process mining, task mining, and observability telemetry, because digital transformation fails when organizations optimize the imagined process rather than the executed one (Object Management Group, 2014; Zerbino et al., 2021; IBM, n.d.; Microsoft, 2024; Celonis, n.d.). [2]
A rigorous digital process analysis lifecycle, therefore, needs eight linked capabilities: discovery, modeling, measurement, analysis, redesign, automation, monitoring, and governance. The primary adaptation to digital transformation is ensuring that each stage is data-aware and execution-aware. Discovery must inventory event data and privacy constraints. Modeling must connect business semantics to system behavior. Measurement must combine process KPIs with technical telemetry. Redesign must include feasibility assessments for automation and control design. Monitoring must remain continuous rather than episodic. Governance must assign ownership not only to process managers but also to data, platform, and change leaders (International Organization for Standardization, n.d.; IBM, n.d.; OpenTelemetry, n.d.; Google Cloud, n.d.). [3]
The strongest practical pattern is layered. Organizations usually begin with a scoping method such as SIPOC or high-level BPMN, use Lean or Six Sigma to frame improvement hypotheses, validate or challenge those hypotheses with process and task mining, redesign the operating model, automate with workflow and low-code platforms where APIs exist, use RPA selectively for legacy or interface-bound work, and instrument the resulting process with logs, metrics, traces, anomaly detection, and conformance monitoring. That layered approach is materially visible in recent public and vendor-documented transformations across government, banking, manufacturing, pharmaceuticals, healthcare, and distribution (American Society for Quality, n.d.; Microsoft, 2025; UiPath, n.d.; SAP, n.d.; ServiceNow, n.d.; Celonis, n.d.). [4]
Foundations of process analysis
Process analysis is the systematic examination of a process to understand how inputs, handoffs, decisions, resources, controls, and outputs produce operational results. IBM [5] defines business process analysis as a detailed, multi-step examination of a process to identify what works, what needs improvement, and how to improve it. At the organizational level, recent BPM research similarly describes business process management as an approach to improving performance and flexibility by managing business processes (IBM, n.d.; Putra & Mahendrawathi, 2024). [1]
Its core objectives are fourfold. First, it seeks transparency, meaning a credible representation of how work actually happens. Second, it seeks performance improvement, usually in the form of lower lead times, lower costs, fewer defects, reduced rework, or higher throughput. Third, it seeks control and conformance, which matters when regulatory, policy, or internal standardization obligations are material. Fourth, in digital transformation, it seeks automation and architectural alignment, meaning that process change is connected to system behavior, data flows, and the realization of transformation value (International Organization for Standardization, n.d.; IBM, n.d.; Microsoft, 2024). [6]
The scope of process analysis can vary substantially. At the narrowest end, it may examine a single task, an approval loop, an exception queue, or a bottleneck. At the broadest end, it may examine end-to-end value streams such as source-to-pay, order-to-cash, claims handling, onboarding, or patient flow. The scope decision is not trivial. Narrow scopes increase tractability but can optimize local work at the expense of total flow. End-to-end scopes produce better transformation decisions but require cross-functional ownership, stronger data integration, and richer governance (Lean Enterprise Institute, n.d.; IBM, n.d.; Celonis, n.d.). [7]
Several concepts are foundational. A process is a structured sequence of actions producing an output for a customer or stakeholder. A value stream is the combined material and information flow required to deliver value. A case is one executed instance of a process. An event log is the timestamped digital record from which process mining reconstructs execution. A variant is a distinct path through the process. Conformance compares executed behavior to intended behavior. Waste, variation, bottlenecks, rework, straight-through processing, and exceptions are analytical lenses rather than isolated metrics. In software-intensive environments, telemetry adds logs, metrics, and traces that reveal runtime behavior beyond the business event log itself (Lean Enterprise Institute, n.d.; American Society for Quality, n.d.; IBM, n.d.; OpenTelemetry, n.d.). [8]
A final distinction is methodologically important. Some approaches are model-driven, relying on workshops, interviews, diagrams, and designed workflows. Others are data-driven, relying on event logs, desktop traces, and telemetry. The strongest contemporary practice links the two rather than choosing between them. This is precisely because BPMN or workshop maps are excellent at making intent explicit, whereas process mining and telemetry are excellent at making execution explicit (Object Management Group [9], 2014; Microsoft, 2024; IBM, n.d.; Celonis, n.d.). [10]
Methodologies and analytical comparison
Traditional and contemporary methodologies should be treated as complementary instruments, not competing ideologies. SIPOC is excellent for scoping. BPMN excels at explicit modeling. Lean and value stream mapping expose waste and flow constraints. Six Sigma excels where variation and defects dominate. Process mining provides evidence of execution from event data. Object-centric process mining is particularly valuable when several business objects interact. Task mining and RPA-driven analysis uncover desktop-level or non-logged work that event logs miss (American Society for Quality, n.d.; Lean Enterprise Institute, n.d.; IBM, n.d.; Microsoft, 2024; UiPath, n.d.; Celonis, n.d.). [11]
| Methodology | Primary analytical lens | Strengths | Weaknesses | Typical use cases | Analytical maturity/implementation complexity |
|---|---|---|---|---|---|
| SIPOC | Boundaries, stakeholders, inputs, and outputs | Rapid scoping, stakeholder alignment, useful before deeper work | Too abstract for redesign by itself | Project framing, chartering, DMAIC Define phase | Very high maturity / Low complexity |
| BPMN | Explicit process logic and handoffs | Common language, executable precision, strong for standardization and governance | Can become diagram-heavy and detached from execution reality | Target operating models, compliance workflows, automation-ready design | High maturity / Medium complexity |
| Lean | Customer value and waste elimination | Strong cultural discipline emphasizes flow and waste | Can underplay statistical variation and system-data realities | Operational excellence, service simplification, backlog removal | Very high maturity / Medium complexity |
| Value stream mapping | End-to-end material and information flow | Good for seeing total flow and delay, ties local work to system outcomes | Often periodic and workshop-driven, less effective without refreshed data | Manufacturing, service journeys, end-to-end redesign | Very high maturity / Low to medium complexity |
| Six Sigma and DMAIC | Variation, defects, capability, root cause | Strong measurement discipline, statistical rigor, durable control mindset | Higher analytic burden, slower when a quick redesign is needed | Quality-sensitive, high-volume, regulated processes | Very high maturity / Medium to high complexity |
| Case-centric process mining | Actual execution paths from event logs | Objective evidence, variant analysis, conformance, automation targeting | Data quality, integration, and case-ID design can be difficult | ERP, CRM, service operations, finance, public-sector workflows | High maturity / High complexity |
| Object-centric process mining | Interactions among multiple business objects | Better for complex, interconnected processes and cross-object dependency analysis | Greater data-model complexity and newer skills are required | Source-to-pay, supply chain, fulfillment, multi-object operations | Growing maturity / High complexity |
| Task mining and RPA-driven analysis | Desktop-level work and interface interactions | Reveals hidden manual work, supports automation sizing in legacy environments | Privacy, employee acceptance, and context interpretation can be challenging | Legacy interfaces, shared mailboxes, swivel-chair work, back-office tasks | Medium maturity / Medium to high complexity |
The table above is an analytical synthesis based on official descriptions by the American Society for Quality [12], Lean Enterprise Institute [13], Microsoft [14], UiPath [15], Celonis [16], IBM [5], and the BPMN specification. The maturity and complexity labels are interpretive ratings derived from standardization depth, skills required, data dependency, and change burden rather than quoted vendor labels (American Society for Quality, n.d.; Lean Enterprise Institute, n.d.; Microsoft, 2024; UiPath, n.d.; Celonis, n.d.; IBM, n.d.; Object Management Group, 2014). [17]
Two analytical cautions follow from this comparison. First, BPMN is a notation, not a self-sufficient improvement method. It becomes analytically powerful only when joined by evidence on measurement, governance, and execution. Second, process mining should not be misread as a complete substitute for human inquiry. Recent reviews show that process mining is powerful for analysis and decision support, but managerial interpretation, process ownership, and contextual understanding remain indispensable, especially when processes are ambiguous, incomplete, or evolving (Object Management Group, 2014; Zerbino et al., 2021; IBM, n.d.). [18]
Adapted lifecycle for digital transformation
An adapted lifecycle for digital transformation differs from a classical improvement cycle in one crucial respect: every phase must be executable in data, not only understandable in workshops. This means that the lifecycle should explicitly connect business semantics, system evidence, automation architecture, and runtime monitoring from the beginning rather than treating technology as a downstream implementation concern (Putra & Mahendrawathi, 2024; Microsoft, 2024; OpenTelemetry, n.d.). [19]
flowchart LR
A[Strategy and governance charter] —> B[Discovery and data inventory]
B —> C[As-is modeling and scoping]
C —> D[Measurement baseline]
D —> E[Multi-method analysis]
E —> F[Redesign and control design]
F —> G[Automation and implementation]
G —> H[Monitoring and observability]
H —> I[Governance review and value realization]
I —> B
E —> J[Quick wins]
J —> H
Discovery. The discovery phase identifies the process boundary, expected business outcomes, relevant stakeholders, systems of record, data owners, and known pain points. In digital transformation, discovery must also inventory available event logs, desktop traces, API access, telemetry coverage, privacy obligations, and automation constraints. If this phase is weak, later mining and automation efforts either fail technically or optimize the wrong process (IBM, n.d.; Microsoft, 2024; UiPath, n.d.; OpenTelemetry, n.d.). [20]
Modeling. Modeling creates the “as-is” and targeted “to-be” representation of the process. High-level scoping can begin with SIPOC, while detailed operational logic is better captured in BPMN, as it is readable by business stakeholders yet precise enough to support software realization. For digital transformation, the model should also identify decision rules, integration points, control points, exception routes, and reusable business objects rather than just ordinary activity sequences (Object Management Group, 2014; American Society for Quality, n.d.; Celonis, n.d.). [21]
Measurement. Measurement establishes the baseline. Traditional measures include lead time, cycle time, first-time-right rate, defect rate, backlog, rework, transport or waiting time, and process capability. Digital transformation expands that baseline to include SLA attainment, straight-through-processing rate, automation potential, manual-touch density, exception frequency, and platform or service telemetry such as latency, error rate, and trace completeness (American Society for Quality, n.d.; IBM, n.d.; OpenTelemetry, n.d.; AWS, n.d.; Google Cloud, n.d.). [22]
Analysis. Analysis should be multi-method. Lean and VSM expose waste and queueing. Six Sigma exposes variation and root causes. Process mining reconstructs actual paths, variants, and conformance gaps. Task mining captures hidden desktop work. Telemetry and anomaly detection reveal runtime drift or hidden system causes. In practice, the strongest analyses triangulate across all of these rather than relying on a single lens (Lean Enterprise Institute, n.d.; American Society for Quality, n.d.; IBM, n.d.; Google Cloud, n.d.; Microsoft Azure, n.d.). [23]
Redesign. Redesign converts insight into target flow. Good redesign does more than “remove steps.” It standardizes variants where possible, simplifies decision logic, clarifies ownership, inserts controls where truly needed, and removes approvals or rekeying that do not add value. In transformation settings, redesign must also determine which changes belong in policy, process logic, integration, and user interface or data architecture (Lean Enterprise Institute, n.d.; IBM, n.d.; SAP, n.d.). [24]
Automation. Automation should follow redesign, not substitute for it. Workflow and low-code platforms are appropriate where APIs, rules, forms, and orchestrated handoffs can be managed centrally. RPA is appropriate where legacy interfaces, desktop steps, or brittle human-system interactions remain unavoidable. The decision criterion is not fashion but technical fit, control needs, and maintainability. In general, API-first orchestration is preferable where available, with RPA reserved for the residual legacy surface (Appian, n.d.; Microsoft, n.d.; Oracle, n.d.; UiPath, n.d.). [25]
Monitoring. Post-implementation monitoring should be continuous. Process mining platforms can monitor conformance, variants, KPIs, and bottlenecks; observability stacks can expose latency, errors, and hidden dependencies; anomaly detection and model monitoring can identify drift in operational or predictive parts of the workflow. Continuous monitoring is the main practical difference between a one-time improvement project and a digital operating model (Microsoft, 2024; AWS, n.d.; Google Cloud, n.d.; Microsoft Azure, n.d.; OpenTelemetry, n.d.). [26]
Governance. Governance closes the loop. Process owners and sponsors should review value realization, unresolved variants, control failures, technical incidents, telemetry blind spots, and the backlog of new automations or redesigns. Governance in this sense is not a committee ritual. It is the structure that prevents process analysis from becoming an isolated project and turns it into an organizational capability (Putra & Mahendrawathi, 2024; SAP, n.d.; Microsoft, 2025). [27]
Digital tools and enabling techniques
Contemporary tooling spans four layers. The first layer is process intelligence and mining. SAP [28] Signavio Process Intelligence is positioned as an AI-powered process mining solution that provides visibility into real process execution and can trigger actions or automations based on process results. Microsoft [14] Power Automate Process Mining emphasizes the discovery, optimization, and identification of automation opportunities. Celonis [16] frames process mining as an objective, real-time view of how processes run and extends this with object-centric process mining for multi-object interactions. UiPath [15] Process Mining and Task Mining explicitly connect macro process discovery to desktop-level evidence (SAP, n.d.; Microsoft, 2024; Celonis, n.d.; UiPath, n.d.). [29]
The second layer is workflow automation and low-code orchestration. Appian [30] presents its platform as a process platform combining low-code, orchestration, automation, intelligence, and a visual drag-and-drop interface. ServiceNow [31] emphasizes no-code and low-code builders, digital workflows, integration, decision tables, and unified flows. Oracle [32] similarly positions Process Automation as low-code orchestration with business rules, integrations, and real-time visibility. SAP Build Process Automation emphasizes low-code and no-code workflow, automation, and broader fusion-team participation. These tools are particularly valuable after analysis has clarified where rules, forms, approvals, integrations, and exception handling should live (Appian, n.d.; ServiceNow, n.d.; Oracle, n.d.; SAP, n.d.). [33]
The third layer is RPA and task-level analysis. UiPath defines RPA as software robots that automate repetitive, rule-based tasks such as data entry and system integration, while Microsoft positions Power Automate as a full-stack automation environment that combines task and process mining, digital process automation, and attended or unattended RPA. In practical terms, RPA is strongest where automation must cross old desktop interfaces, thin clients, email-heavy work, or systems without usable APIs. Task mining serves as a discovery mechanism that quantifies how much of that work is genuinely automatable and where redesign is needed before a bot should be built (UiPath, n.d.; Microsoft, n.d.). [34]
The fourth layer is analytics, anomaly detection, and observability. Microsoft Azure’s Metrics Advisor and Anomaly Detector support multidimensional monitoring, root-cause diagnosis, and time-series anomaly detection. Google Cloud [35] provides BigQuery anomaly detection and Vertex AI Model Monitoring to track deviations, drift, and thresholds, and to trigger alerts. Amazon Web Services [36] CloudWatch connects metrics, logs, and traces and explicitly supports OpenTelemetry. OpenTelemetry itself provides the instrumentation framework and common conventions for traces, metrics, and logs. For digital transformation, these tools matter because the process no longer ends at the business event log. Runtime reliability, machine-learning drift, and platform performance can materially alter the business process outcome (Microsoft Azure, n.d.; Google Cloud, n.d.; AWS, n.d.; OpenTelemetry, n.d.). [37]
A practical selection rule follows. If the organization’s problem is “we do not understand the real flow,” start with process mining. If the problem is “we understand the flow, but handoffs and rules remain manual,” move toward orchestration and low-code. If the problem is “the critical work sits in legacy interfaces,” add RPA. If the problem is “the automated process is now running but not reliably,” instrument it with logs, metrics, traces, anomaly detection, and drift monitoring. In mature environments, all four layers coexist (Microsoft, 2024; Appian, n.d.; UiPath, n.d.; AWS, n.d.; Google Cloud, n.d.). [38]
Case studies in digital transformation
Municipality of Rotterdam [39], public sector. Rotterdam used Power Automate Process Mining to analyze more than 30 processes across multiple systems, from citizen feedback routing to grants and election monitoring. One grants-related analysis that had previously taken all day could be completed in 10 minutes after the shift. The deeper lesson is not merely speed. The municipality also found that broader access to process data increased employee motivation to act on findings, which is a useful reminder that democratized visibility can be a change mechanism in its own right (Microsoft, 2025). [40]
İşbank [41], banking. UiPath documents that İşbank mined 26 processes, saved 116,000 hours, optimized 61% of cases by removing unnecessary approval steps, and improved 65% of cases against SLA expectations. The notable lesson is methodological: process mining created the evidence needed to challenge inherited approval logic. In digital transformation, many “controls” are actually historical frictions whose marginal risk-reduction value is lower than their operational cost (UiPath, n.d.). [42]
Louisiana-Pacific Corporation [43], manufacturing. SAP reports that Louisiana-Pacific identified an estimated US$24 million in value potential across more than 15 use cases after just 48 hours of using the solution. Equally important, the company emphasized that process insights should not be framed as criticism of individuals but as empowerment for process owners, and that process ownership is critical to successful transformation. This case sharply illustrates that large-value cases can emerge quickly from data-driven process discovery, but realization depends on process ownership and change management rather than on analytics alone (SAP, n.d.). [44]
Ingram Micro [45], distribution and order-to-cash. Celonis reports that Ingram Micro eliminated 50,000 labor hours by identifying optimization and automation opportunities in the order-to-cash process using a single metric. The lesson here is strategic prioritization. Process analysis does not need to start with an enterprise-wide overhaul. Focusing on a load-bearing metric in a revenue-critical process can quickly deliver demonstrable value and create the conditions for scale (Celonis, n.d.). [46]
AstraZeneca [47], life sciences. ServiceNow reports that AstraZeneca now routes 60,000 laboratory requests through one system, saves more than 30,000 hours annually, projects more than 90,000 hours of savings for hiring managers through onboarding transformation, and reduces tasks that previously took 30 minutes to seconds. This case exemplifies a broader transformation principle: where scientific or knowledge-intensive organizations depend on experts, process analysis should be used to remove administrative burden from high-value labor rather than merely to cut transaction cost (ServiceNow, n.d.). [48]
Gold Coast Health [49], healthcare. UiPath reports that automation reduced the time to access clinical images in patient records from 19 hours to 15 minutes and made the flow available 24/7. The practical lesson is that healthcare value is frequently created not only through “front-end” patient-facing redesign but through back-office process reliability, metadata capture, and reduced administrative delay in the clinical context (UiPath, n.d.). [50]
Across these cases, four recurring lessons appear. First, measurable value often comes from variant reduction, approval simplification, and better information flow rather than from exotic AI alone. Second, process analysis is most credible when it links operational metrics to stakeholder outcomes. Third, adoption strengthens when the processed data is shareable and understandable to business users. Fourth, governance and change ownership are decisive after the first wave of insight is generated (Microsoft, 2025; SAP, n.d.; UiPath, n.d.; Celonis, n.d.; ServiceNow, n.d.). [51]
Governance, change management, and measurement
Best practice begins with business-outcome-first framing. The process should be analyzed because organizational outcomes matter, including faster case resolution, higher first-time-right rates, better working capital, lower claim delays, stronger compliance, and reduced employee effort. From there, scope discipline is essential. Start with an end-to-end lens, but identify the few metrics and variants that matter most. Use multi-method triangulation, combining stakeholder maps, explicit models, and execution evidence—design for exception handling and controls, not just happy paths. Keep the redesign API-first where possible and bot-assisted where necessary. Finally, make the post-go-live process observable, not just deployed (IBM, n.d.; Microsoft, n.d.; Oracle, n.d.; OpenTelemetry, n.d.; AWS, n.d.). [52]
Common pitfalls are equally consistent. Organizations frequently model the process they believe exists rather than the one their systems and people actually execute. They treat BPMN diagrams as deliverables rather than as decision artifacts. They automate unstable workarounds with bots, thereby scaling poor design. They neglect data quality and case-identification issues in event logs. They ignore manual work occurring outside logged systems. They underinvest in ownership and change leadership after the analytics phase. Moreover, they track only implementation output, such as the number of bots or diagrams produced, rather than realized process outcomes (IBM, n.d.; Zerbino et al., 2021; UiPath, n.d.; SAP, n.d.). [53]
Change management should be built into the method rather than attached afterward. Recent customer evidence from Louisiana-Pacific shows the importance of communicating that process insight is meant to empower people, not criticize them, and that process ownership must be explicit. Rotterdam similarly shows that easier access to process data can widen participation and accelerate local problem-solving. These patterns support a governance model in which an executive sponsor, a process owner, a data owner, a platform or automation owner, and a change lead share accountability over a common transformation backlog (SAP, n.d.; Microsoft, 2025). [54]
A useful KPI framework for analyzing digital transformation processes has five layers. Flow KPIs track lead time, cycle time, throughput, waiting time, touch time, backlog, and SLA compliance. Quality and control KPIs track first-time-right, defect rate, rework rate, exception rate, audit findings, and conformance deviations. Automation KPIs track straight-through-processing rate, automation coverage, bot success rate, human exception ratio, and manual step reduction. Experience KPIs track customer effort, satisfaction, abandonment, and employee effort or time reclaimed. Runtime and AI KPIs track latency, error rate, trace completeness, anomaly frequency, data drift, and alert precision. The point is not to maximize the number of KPIs but to tie each layer to a specific decision mechanism and owner (IBM, n.d.; OpenTelemetry, n.d.; AWS, n.d.; Microsoft Azure, n.d.; Google Cloud, n.d.). [55]
For mature programs, the measurement framework should be staged. A baseline stage records the current-state process and telemetry measures. A diagnostic stage links those measures to root causes, variants, drift, and technical dependencies. A value-realization stage measures the delta after redesign and automation. A control stage governs sustainability through thresholds, alerts, conformance checks, and recurring owner reviews. That is the progression by which process analysis becomes a durable transformation capability rather than a one-off improvement project (American Society for Quality, n.d.; IBM, n.d.; Google Cloud, n.d.; OpenTelemetry, n.d.). [56]
References
American Society for Quality. (n.d.). DMAIC process: Define, measure, analyze, improve, control.
American Society for Quality. (n.d.). Quality glossary of terms, acronyms & definitions.
American Society for Quality. (n.d.). Six Sigma.
American Society for Quality. (n.d.). Value stream mapping tutorial: What is VSM?
Appian. (n.d.). Appian low-code platform.
Appian. (n.d.). Business process automation, explained.
AWS. (n.d.). Amazon CloudWatch.
Celonis. (n.d.). Getting started with object-centric process mining.
Celonis. (n.d.). Ingram Micro + Celonis | Process Mining.
Celonis. (n.d.). What is process mining? Definition and benefits.
Google Cloud. (n.d.). Anomaly detection overview.
Google Cloud. (n.d.). Introduction to Vertex AI Model Monitoring.
Google Cloud. (n.d.). Observability: Cloud monitoring and logging.
IBM. (n.d.). What is business process analysis?
IBM. (n.d.). What is process mining?
International Organization for Standardization. (n.d.). Quality management principles.
Lean Enterprise Institute. (n.d.). Lean thinking and practice.
Lean Enterprise Institute. (n.d.). Value stream mapping overview.
Microsoft. (2024). Overview of process mining in Power Automate.
Microsoft. (2025). The Municipality of Rotterdam optimizes services for citizens using insights from Power Automate Process Mining.
Microsoft. (n.d.). Power Automate.
Microsoft Azure. (n.d.). AI Anomaly Detector.
Microsoft Azure. (n.d.). Pricing: Azure AI Metrics Advisor.
Object Management Group. (2014). About the Business Process Model and Notation Specification Version 2.0.2.
OpenTelemetry. (n.d.). What is OpenTelemetry?
Oracle. (n.d.). Application integration and data integration.
Putra, H., & Mahendrawathi, E. R. (2024). The role of business process management in digital innovation and digital transformation: A systematic literature review. Procedia Computer Science, 234, 829–836.
SAP. (n.d.). Business process automation software | Low code.
SAP. (n.d.). LP: Building a new path for process professionals to accelerate business transformation and roll out change.
SAP. (n.d.). SAP Signavio Process Intelligence.
ServiceNow. (n.d.). AstraZeneca.
ServiceNow. (n.d.). Workflow automation.
UiPath. (n.d.). From data to action: İşbank’s journey of operational excellence with process mining.
UiPath. (n.d.). Gold Coast Health uses automation to deliver real-time patient care.
UiPath. (n.d.). Starting a Task Mining project from Process Mining.
UiPath. (n.d.). What is robotic process automation?
Zerbino, P., Stefanini, A., & Aloini, D. (2021). Process science in action: A literature review on process mining in business management. Technological Forecasting and Social Change, 172.
[1] [13] [20] [28] [41] [45] https://www.ibm.com/think/topics/business-process-analysis
https://www.ibm.com/think/topics/business-process-analysis
[2] [10] [18] [21] https://www.omg.org/spec/BPMN/2.0.2/About-BPMN
https://www.omg.org/spec/BPMN/2.0.2/About-BPMN
[3] [6] [39] https://www.iso.org/iso/pub100080.pdf
https://www.iso.org/iso/pub100080.pdf
[4] [11] [17] https://asq.org/quality-resources/quality-glossary?srsltid=AfmBOorvpMgJYmZ7eGCvB2WZz8FEzpql1N1OoovxG8qNwQ3NiaeDL7v7
[5] [48] https://www.servicenow.com/customers/astrazeneca.html
https://www.servicenow.com/customers/astrazeneca.html
[7] [8] [23] [30] https://www.lean.org/lexicon-terms/value-stream-mapping/
https://www.lean.org/lexicon-terms/value-stream-mapping/
[9] [26] [38] https://learn.microsoft.com/en-us/power-automate/process-mining-overview
https://learn.microsoft.com/en-us/power-automate/process-mining-overview
[12] [15] [29] [35] https://support.sap.com/en/product/onboarding-resource-center/sap-signavio/pint.html
https://support.sap.com/en/product/onboarding-resource-center/sap-signavio/pint.html
[14] [16] [34] https://www.uipath.com/rpa/robotic-process-automation
https://www.uipath.com/rpa/robotic-process-automation
[19] [27] [49] https://www.sciencedirect.com/science/article/pii/S1877050924004277
https://www.sciencedirect.com/science/article/pii/S1877050924004277
[22] [31] [43] [56] https://asq.org/quality-resources/dmaic?srsltid=AfmBOorXp5MU0Qrxg096evYog5J16f8wYyRaXgA4Lky6z\_GN-uuJ8NqG
[24] https://www.lean.org/lexicon-terms/lean-thinking-and-practice/
https://www.lean.org/lexicon-terms/lean-thinking-and-practice/
[25] [33] [47] https://appian.com/products/platform/low-code
https://appian.com/products/platform/low-code
[32] [40] [51] The Municipality of Rotterdam optimizes services for citizens using insights from Power Automate Process Mining | Microsoft Customer Stories
https://www.microsoft.com/en/customers/story/25307-gemeente-rotterdam-power-automate
[36] [44] [54] LP: Building a new path for process professionals to accelerate business transformation and roll out change
https://www.sap.com/hk/asset/dynamic/2025/04/ec325c1a-ff7e-0010-bca6-c68f7e60039b.html
[37] Pricing - Azure AI Metrics Advisor | Microsoft Azure
https://azure.microsoft.com/en-us/pricing/details/metrics-advisor/
[42] https://www.uipath.com/resources/automation-case-studies/isbank-journey-of-operational-excellence-with-process-mining
[46] https://www.celonis.com/solutions/stories/ingram-micro-celosphere-keynote-process-mining
https://www.celonis.com/solutions/stories/ingram-micro-celosphere-keynote-process-mining
[50] https://www.uipath.com/resources/automation-case-studies/gold-coast-health-uses-automation-to-deliver-real-time-patient-care
[52] [55] https://www.ibm.com/think/insights/process-mining-benefits
https://www.ibm.com/think/insights/process-mining-benefits