Tools and methods for AI-enabled process transformation

Building capability to understand, shape, and scale operational systems over time.

ProcessKAS is a common home for tools, thinking, and methods that address how operational systems actually behave—through data, decisions, constraints, and feedback. Built on a straightforward framework: Know → Act → Scale.

Know → Act → Scale

Transformation is about building disciplined capability: sensing how work behaves over time, acting on insight, and embedding change at scale. The focus is on how work behaves over time, not idealised process maps.

Know

Understanding how operational systems actually behave. Event data, process mining, decision logic, and semantic models that capture what's happening—variation, rework, feedback, constraints—not static flowcharts.

Act

Designing and implementing interventions. Redesign, automation, controls, decision rules. Changes grounded in how systems actually work, not how they're documented.

Scale

Embedding capability and sustaining change. Standards, governance, decision rights, and the systems that allow organisations to continuously adapt and learn.

Work & Tools

ProcessKAS brings together tools, prototypes, and experiments built around a consistent way of understanding and shaping operational systems. These artefacts vary in maturity, but they are treated as first-class expressions of the same underlying approach.

Active

Synthetic Event Log Generator

Configurable generators for realistic process event logs. Supports process mining research, training, and solution design.

Python PM4Py
In Development

Process Knowledge Graph

Semantic and knowledge-graph-based representations of processes and decisions. Built for querying, reasoning, and generating artifacts from declarative models.

Python RDF SPARQL
Exploratory

Open-Source Process Mining Pipeline

Minimal end-to-end implementation showing how raw operational data becomes event logs and process models—using SQL, Python, and PM4Py.

Python SQL PM4Py
Exploratory

NFL Process Mining Example

Worked example treating an event log as a designed dataset, using public sports data to make process mining concepts concrete and testable.

Python Pandas
Exploratory

Process Intelligence Data Assessment

Utilities for exploring data readiness for process analytics—assessing structure, quality, and constraints before transformation work begins.

Python SQL
Exploratory

Process Semantic Layer

AI-adjacent tools for reasoning about operational systems. Exploring whether lightweight business concepts can make retrieval-based AI more explainable and auditable.

Python LLM
View GitHub Profile

Not all work is surfaced here. Some projects exist as early-stage or transitional artefacts.

Applying the work

The methods and tools on this site reflect a particular approach to process transformation—one grounded in systems thinking, operational discipline, and capability building. Six principles guide that approach.

Capability over projects

Lasting change comes from building organisational capability, not from delivering isolated initiatives. Tools, insights, and interventions only matter if they strengthen how an organisation continuously understands and manages its own processes.

The focus is on how work actually behaves over time — through variation, feedback, rework, and decision-making — rather than on static process maps or idealised models.

Process insight, including process mining, is a diagnostic capability. It helps explain what is happening and why. It does not, on its own, create transformation.

Automation and AI amplify whatever structure already exists. Without clear process semantics, constraints, and decision logic, AI systems scale inconsistency rather than capability.

Scaling change requires clear ownership, standards, and decision rights. Governance is not overhead — it is what allows learning and improvement to persist beyond individual efforts.

Process transformation is an ongoing system of sensing, acting, and learning. The goal is not a finished 'to-be', but a disciplined ability to adapt as conditions change.

About

ProcessKAS is grounded in a specific worldview: that operational systems are observable, shapeable, and AI-ready only when their behaviour is understood through data, decisions, constraints, and feedback—not static diagrams.

The work focuses on AI-enabled process transformation: how organisations can design operational systems that are explainable, adaptable, and capable of embedding both human and machine decision-making. This isn't process mining for its own sake—it's one diagnostic technique among many. The centre of gravity is process intelligence and the broader discipline of shaping operational systems over time.

This isn't exploratory research without application. It's architectural thinking grounded in real operational problems: how do you build systems that can be reasoned about, improved continuously, and made ready for automation and AI—without depending on fragile, vendor-locked abstractions?

The Know → Act → Scale framework emerged from years of work where diagnostic insight without intervention goes nowhere, and intervention without governance fades. It's a way of thinking that unifies tool development, method design, and real-world capability building.

Connect

For questions about the methods, tools, or to discuss collaboration opportunities.

Connect on GitHub Connect on LinkedIn