Celonis — Process Mining & Automation UX

Context

Celonis is the market leader in process mining and process automation — enterprise software that helps organisations analyse and optimise their operational processes. The product surface is inherently complex: it works with large datasets, deep process models, and diverse user roles ranging from business analysts to technical administrators.

I contributed as Senior UX Designer on core journeys for process mining and automation use cases — working across multiple iterations with product managers and engineering teams.

Challenge

Data-heavy enterprise tools can overwhelm users. The challenge was to reduce cognitive load without removing the depth and precision that power users depend on. In automation tools especially, incomplete or ambiguous states — tasks in flight, partially complete processes, error conditions — can erode user trust fast.

An additional challenge: multiple teams own different parts of the product surface, which means consistency and shared patterns matter more, not less.

What I did

• Designed and iterated interaction patterns for complex, stateful workflows — tasks, automation outcomes, error states, and empty states
• Built high-fidelity interactive prototypes for stakeholder alignment and to de-risk engineering feasibility before build
• Led multiple rounds of usability testing and translated findings into clear, prioritised product requirements
• Collaborated closely with product managers and engineers to ship incremental improvements across several release cycles
• Contributed patterns to the shared design system to reduce duplication across teams

Process

Selected visuals

Outcome

• Cleaner workflows with clearer task and state handling — reduced screen complexity without removing necessary depth
• Faster task completion for frequent operations through better information hierarchy and affordances
• Smoother adoption across enterprise user groups via predictable, consistent patterns
• Improved cross-team alignment through prototypes as shared decision-making artefacts

What I learned

• When everything is important, nothing is — hierarchy is the product in data-heavy tools
• State modelling (empty, loading, partial, error, success) prevents "mystery meat" automation experiences that erode trust
• Prototypes are the fastest alignment tool when multiple teams own different parts of the same flow
• Enterprise UX improvement is almost always incremental — the skill is identifying the highest-leverage changes within real constraints