Process Mining Software

Process mining software for operations and transformation analytics

Process mining software reverse-engineers process models from system event logs - the as-is emerges from what the systems show actually happened, not from what an analyst thinks happens. This page covers the mining workflow as a BPMN process map and how mining pairs with discovery and BPM.

By Jack Finnegan · Updated 21 May 2026

Model what mining surfaces Process discovery

What it is

What process mining software actually is

Process mining software derives process models from system event logs. The technique was pioneered by Wil van der Aalst at Eindhoven University of Technology in the late 1990s (the term was coined in his 1999 research proposal) and has since matured into a commercial category. The input is an event log - typically (case ID, activity, timestamp) tuples extracted from ERP, CRM, or workflow systems. The outputs are a discovered model showing what the systems recorded actually happened, conformance analysis comparing actual to documented behaviour, and performance analysis on cycle time, throughput, and rework.

Major platforms include Celonis (the category leader, founded in 2011), UiPath Process Mining (built on the 2019 ProcessGold acquisition), Apromore (open-source-rooted, acquired by Salesforce in November 2025), ARIS Process Mining (part of the ARIS portfolio, which became a standalone business in January 2025 following the Software AG restructure), and IBM Process Mining (built on the 2021 myInvenio acquisition). SAP integrated process mining via the 2021 Signavio acquisition; Microsoft entered via the 2022 Minit acquisition. The category increasingly overlaps with task mining (which observes user actions rather than system events) and with traditional BPM tooling that adds mining capability.

The problem today

Mining shows the deviation, then nobody redesigns the process

The pattern: a firm deploys process mining, identifies a 30% deviation from the documented process, presents the heat-map at the executive committee, and... nothing happens. Mining surfaced the problem but the team did not have the BPMN process redesign capability to act on it. Six months later the dashboard is still showing the deviation; the firm wonders why the mining licence costs were not justified by improvement outcomes.

The fix is pairing mining with discovery and improvement. Mining identifies where reality diverges from the documented model; discovery captures the workarounds that explain why; improvement designs the fix; BPM-style modelling produces the new as-is. Without all four steps, mining is observation without action.

Four pillars

Four pieces of a working mining engagement

Event-log quality

Mining is only as good as the log. Missing timestamps, duplicate case IDs, inconsistent activity names all degrade the model.

Discovery vs conformance

Use discovery when there is no documented model; use conformance when there is and you suspect drift.

Performance analysis

Cycle time, throughput, rework. The numbers are what justify acting on the deviations.

Hand-off to improvement

Mining without improvement is observation. The mined model has to feed redesign, automation, or compliance remediation to be worth the licence cost.

Process Map

A mining engagement as a process map

The canonical flow - scope, extract, discover, log-quality check, conformance, performance, recommend, hand off.

Open in editor

A process mining engagement as a process map

A process mining engagement rendered as a BPMN 2.0 process. Connect to source systems, extract event logs, discover the as-is from data, run conformance and performance analysis, surface deviations, and recommend improvements.

Identify the process and source systems that hold the event log.
Extract and prepare event logs (case ID, activity, timestamp, attributes).
Run process discovery to derive the as-is model from the log.
Check log completeness: if gaps in case coverage or attributes, loop back to extract more data; otherwise proceed.
Run conformance checking against the expected model (if one exists).
Analyse performance - cycle time, throughput, rework loops.
Recommend improvements based on the conformance and performance findings.
Hand off to the improvement or modernization programme.

What this diagram shows: Once the scope is agreed, event-log extraction comes first. The discover task derives the model from the log; the log-quality gateway routes incomplete logs back to extraction for more data and forwards complete logs to conformance checking. Conformance compares the discovered model against expected behaviour; performance analysis quantifies cycle time and rework. The Recommend Improvements task turns the conformance and performance findings into a remediation list, which the hand-off task forwards to the improvement or modernization programme.

FAQ

Frequently asked questions

What is process mining software?

Process mining software derives process models from system event logs and analyses conformance and performance. Major platforms include Celonis, UiPath Process Mining, Apromore (now part of Salesforce), ARIS Process Mining, and IBM Process Mining.

How is process mining different from process discovery?

Process mining is one specific technique (model from event logs). Process discovery is the broader discipline that also includes interviews, observation, and document review. Mining is an input to discovery, not a replacement - it cannot capture manual steps that happen outside systems.

What is task mining?

Task mining observes user actions at the client (mouse clicks, keystrokes, screen recordings) rather than system events at the server. It captures process detail that does not reach the event log (e.g. manual copy-paste between systems). Most modern mining platforms combine task mining with event mining.

Does process mining replace BPM software?

No. Mining produces the as-is; BPM software produces the documented model and (where applicable) executes the process. The two are complementary - mining feeds the BPM lifecycle by showing where the documented model and the lived process have diverged.

Does BA Copilot integrate with process mining tools?

Indirectly. BA Copilot is the modelling layer - it produces and refines the BPMN diagrams that mining outputs hand off to. Celonis, UiPath, and ARIS export their discovered models in BPMN or related notations; BA Copilot edits, refines, and produces the to-be designs the improvement programme runs against.

Jack Finnegan, Founder & CEO, BA Copilot

From the founder

14 Years in BPMN

I'm Jack Finnegan. I've spent fourteen years working hands-on with BPMN, as an analyst, an engineer, and a product director, where I felt every sharp edge of legacy business process platforms.

BA Copilot is the platform I wanted on every one of these projects: AI-first process management, which treats BPMN as a first-class output rather than an export afterthought.

Jack Finnegan on LinkedIn

Turn mining output into action

Open the mined process in BA Copilot, refine the BPMN, design the to-be, and hand off to the improvement programme - the step mining alone does not close.