Celonis, AI-Driven Process Mining

Celonis Starter Kits are pre-built onboarding experiences designed to help enterprise customers quickly realize value from Celonis’ process mining platform. Over time, however, these kits had become complex and difficult to configure—slowing onboarding and increasing reliance on hands-on support.

Within the constraints of Celonis’ low-code Studio platform, the challenge wasn’t just to “improve UX.” It was to reduce friction today while defining a clearer, more scalable future.

Over six months, I helped lead a dual-track approach:

Phase 1 (Near Star): Improve usability within platform constraints to accelerate time to value

Phase 2 (North Star): Define a forward-looking vision that reimagined onboarding as a more intuitive, AI-supported journey

The result paired immediate usability gains with a compelling product vision that aligned stakeholders, influenced roadmap thinking, and elevated how Celonis imagines its future experience.

Celonis Starter Kits are pre-built onboarding experiences that help enterprise customers quickly get value from the Celonis process mining platform. They bundle dashboards, validation tools, and workflows for business functions like Accounts Receivable and Procurement.

Over time, Starter Kits had become complex and difficult to configure. Users struggled to understand where to begin, how to validate data, and how to move from insight to action. Onboarding slowed. Support dependency increased.

Compounding the issue, the experience was built inside Celonis Studio—a low-code environment where even small UX improvements required careful tradeoffs.

The challenge was clear:

Improve what exists today while defining a smarter, more intuitive future.

Approach

We structured the engagement in two phases to balance practicality with long-term vision.

Phase 1 — Near Star (Foundation + Learning)

The first phase focused on improving usability within existing constraints.

I helped:

Audit the Starter Kit experience to identify friction and hierarchy breakdowns

Review user sessions and internal demos to surface common stall points

Refine navigation, information architecture, and visual consistency

Clarify next steps and reduce cognitive overload

This work stabilized the experience and clarified which problems were incremental—and which required a more fundamental rethink.

Phase 2: North Star (Primary Focus)

With a deeper understanding of user friction and platform limits, we shifted from “What can we tweak?” to:

“What would onboarding feel like if it were genuinely intuitive?”

In this phase, we:

Mapped an ideal end-to-end onboarding journey

Identified where AI could reduce manual effort and build confidence

Designed a guided experience that reframed setup as exploration rather than configuration

Key concepts included:

Sample-data onboarding, allowing users to explore value before uploading real data

AI-supported insights, pairing conversational guidance with data visualizations to help users move from insight to action

Rather than replacing user judgment, AI was positioned as a confidence-building partner.

Outcome

The dual-track approach delivered impact at multiple levels:

Near-term usability improvements reduced friction and supported faster time to value

A clear North Star vision aligned teams around a more intuitive, AI-supported future

Together, the work balanced constraint-driven design with forward-looking product thinking—improving the present while shaping the roadmap.

Celonis is an enterprise process mining platform used by large organizations to analyze and optimize complex operations. One of its primary entry points is through Starter Kits—pre-built packages of dashboards, validation tools, and workflows designed for business functions like Accounts Payable, Accounts Receivable, and Procurement.

Starter Kits are intended to accelerate time to value.

Over time, however, they had become increasingly complex:

Setup required significant guidance

The UI felt noisy and fragmented

Insights required heavy manual interpretation

Users struggled to understand where to start

At the same time, the experience was built inside Celonis Studio—a low-code, drag-and-drop environment that limited how much the UX could evolve in the short term.

The team needed to improve the current experience without overcommitting to premature build decisions—and define a clearer long-term direction.

Problem

Starter Kits were meant to accelerate value. Instead, they were slowing customers down.

Onboarding required internal expertise

Navigation lacked clear hierarchy

Insights were surfaced, but not guided

Platform constraints made even small improvements costly

How might we reduce friction today while pointing toward a smarter, more intuitive future experience?

Goals

Success wasn’t about pixel perfection. It was about clarity and alignment.

We aimed to:

Reduce friction in Starter Kit onboarding

Help users move from insight to action faster

Work within Studio constraints without lowering UX quality

Define a North Star to guide future investment

Explore where AI could build trust and confidence

Strategic Approach: Two Phases, One Narrative

We structured the engagement intentionally:

Phase 1 (Near Star): Improve what exists and learn where real constraints lie

Phase 2 (North Star): Step beyond those constraints to define an ideal future experience

This sequencing ensured immediate value while building strategic clarity.

Phase 1 — Near Star (Foundation + Learning)

Discovery & Audit

We conducted a deep product audit:

Mapped Starter Kit variations and flows

Reviewed user sessions and demo recordings

Annotated friction points, IA breakdowns, and usability gaps

Identified where users stalled or required intervention

We also reviewed competitor patterns around guided setup and automation.

Near Star Improvements

Within Studio constraints, we focused on pragmatic refinements:

Re-established visual hierarchy

Improved navigation clarity

Reduced noise and unnecessary elements

Made next steps explicit

This work stabilized onboarding and surfaced which challenges were incremental versus systemic.

Phase 2 — North Star (Primary Focus)

With clarity around constraints, we stepped back.

Instead of asking “What can we tweak?” we explored:

“What would onboarding feel like if it were intuitive from the start?”

North Star Framing

The vision focused on helping users quickly understand:

What the system is doing

Why insights matter

What to do next

We reframed onboarding as a guided, exploratory journey designed to build confidence—not demand expertise.

Key North Star Concepts

1. Sample-Data Onboarding

Pre-populated dashboards with sample data

Allowed exploration before real data upload

Reduced fear of “doing it wrong”

Built early confidence

2. AI-Supported Insights

Integrated conversational guidance to validate findings

Paired AI outputs with visual evidence

Reduced manual analysis

Accelerated movement from insight to action

AI was positioned as a collaborator—not a replacement for judgment.

Prototyping & Handoff

To bring the North Star vision to life, we:

Built an interactive prototype demonstrating key flows

Used micro-animations to show transitions and system behavior

Documented intent, assumptions, and open questions

Supported handoff and alignment across product, design, and leadership

The prototype became a shared artifact for conversation, alignment, and future planning.

Outcome

This two-phase approach delivered impact at multiple levels:

Near Star: Reduced friction in the existing Starter Kit experience and supported faster time to value

North Star: Helped teams align around a clearer, more intuitive future experience

The work elevated the perceived maturity of the product while leaving room for future validation and iteration.

Learnings

Designing within constraints builds clarity, but knowing when to step beyond them is equally important

North Star concepts can be effective tools for alignment when paired with grounded Near Star improvements

AI is most effective when it supports user confidence rather than replacing judgment

Challenges

Rigid low-code tooling made even small UX changes costly

A noisy, fragmented UI required re-establishing hierarchy from the ground up

Limited access to end users shaped how feedback was gathered and evaluated

Wins

Helped shape a dual-track Near Star + North Star approach

Used design concepts to support internal alignment and shared understanding

Delivered thoughtful UX despite tight technical constraints

What I’d Do Next

With more time, I’d validate the North Star concepts with customers to better understand where AI builds the most trust and confidence.