Celonis · Product Design
January, 2025Annotation Builder
Designed an AI-powered production tool for data scientists that enriches free-text data. Made “ease of use” an enterprise product's exit criterion and, starting from a zero state, made it possible to reach thousands of users in 12 weeks.
ARR Influenced, 2025
$230M+
Time to Build
12 Weeks
Customer Rating
4.08 / 5
Logos Reached
338
Task Completion Rate
88%
vs 67.2% platform avgTime on Task
18 min
vs 244 min platform avg
Celonis customers use process mining to derive deep insights from operational data across complex workflows such as supply chain, order management, and manufacturing. A key challenge is the volume of raw, unstructured free-text data, which often requires significant effort from multiple analysts to process.
- I owned the end-to-end design process, from framing the ideal solution with Product Management, through engineering handoff, defining UX success criteria, and tracking metrics. I also shaped the core prompt interactions, which Celonis introduced to the platform for the first time.
- The main value proposition was ease of use: most feature and scaling decisions were driven by user experience, marking a significant mindset shift for an enterprise data company.
- At its core, Annotation Builder enables users to enrich free-text data through a low-barrier, progressively disclosed experience, avoiding overwhelm, especially when compared to traditional machine learning configurations or automation setups.
- As a result, users can make their data actionable within minutes, rather than spending days on configuration and analysis.
Manually processing large volumes of free-text data through machine learning algorithms or predefined automations is a significant bottleneck, because human judgment is required at each step. This demands substantial manual effort, delays customers' time to value, and increases operational burn rate.
- Costly inaccuracies: Manual processes mean human error, leading to substantial revenue impacts for an average Celonis customer, ranging between $350,000 and $10 million+ annually.
- Strategic AI Leadership: Celonis pioneered agentic AI within process mining. This initiative not only solves immediate customer pain points but also enabled us to integrate with Microsoft Azure & OpenAI as an agentic partner, and solidifies our leadership in AI-driven process intelligence alongside companies like Palantir and SAP.
Click to see Annotation Builder in action.
Research Approach
Conducted contextual inquiry and stakeholder interviews to ground design decisions in real user workflows and enterprise data patterns.
Problem Definitions, Solution Proposals & Hypotheses
Synthesised research into clear problem statements, explored multiple solution directions, and formulated testable hypotheses before committing to a design path.
UI Craft & Design Decisions
Shaped interaction patterns, progressive disclosure mechanics, and prompt UX, introducing conversational AI interactions to the Celonis platform for the first time.
Validation Process
Ran moderated usability studies and unmoderated concept tests to validate hypotheses, then iterated based on behavioural signals and qualitative feedback.
Cross-Functional Collaboration
Partnered closely with Product Management and Engineering throughout, from framing to handoff, ensuring design intent translated into the shipped product.
Since its general availability release in November 2024, Annotation Builder has been rolled out to 338 Celonis customers and has become the top sales-driving feature for the current fiscal year. We continue to receive overwhelmingly positive feedback on its:
- Ease of use
- Accuracy of execution
- Short time to value
This demonstrates how strategic UX input can be a significant revenue driver for enterprise applications, directly challenging the common misconception that enterprise users don't need simple interfaces.
“Based on your initial expectations, how well did the Annotation Builder meet your expectations?”
39 individual customer participants — post-release, November 2024
Customer sentiment: Examples from 2025
“The configuration is well structured and easy to implement. Ticket routing took only one day from development to production.”
“AI is not perfect yet, but Annotation Builder works great for us. Free-text categorization is especially promising.”
“It is a powerful tool that performs exceptionally well. The configuration process is straightforward, and the results align with expectations. During the AI Labs, it took one day to reach productivity.”
“I am very positively surprised by how Annotation Builder solves complex tasks through reasoning. It aligns with our customer problems.”
“I was surprised by how well it captures fuzzy business rules and produces well-defined decisions. Prompt-based configuration makes it easy to build.”