Data Prep for AI-Driven Enterprises

This initiative seamlessly integrated advanced data preparation capabilities into a unified platform. The integration empowered data scientists of all skill levels to efficiently collect, ingest, combine, cleanse, and transform data within a single environment.

Project Overview

Contribution

Principle UX Designer, Interaction Design, Prototyping

Target Users

Data Scientists, Data Engineers

Products

Enterprise Data Management

Timeline

4 months

Impact

Integrated Data Prep improved overall productivity and accelerated AI development cycles across teams.

40%

Faster Workflow Completion

100%

No-Code Interface Adoption

12x

Faster Insights

Solution

Seamlessly integrating Paxata's data preparation functionalities into DataRobot to create a cohesive and unified experience.

UI-Driven Data Prep

Intuitive interface for data scientists of all skill levels, enabling self-service data preparation without coding.

SQL-Based Transformation

Advanced capabilities for experienced users, providing powerful SQL-based data transformation tools.

Unified Workflow Experience

Data preparation is seamlessly integrated into a Unified Workflow Experience, allowing users to complete data prep tasks within a single page application for a cohesive and consistent experience.

Workflow Experience

Users have access to tools for collecting, cleansing, and transforming data, allowing them to build end-to-end machine learning pipelines without switching between platforms.

Data Transformation

Integrated data prep removes bottlenecks, enabling customers to scale ML use cases and deliver results faster.

Recipe Builder

Interactive Data Pipeline Builder

Testing Alternate Design Patterns: I explored and tested two distinct UI patterns, evaluating users' time on task, task success, and perceived ease of use. The graph editor approach offers an intuitive and powerful visual approach that simplifies the learning curve for new data scientists while boosting efficiency for experienced users.

Final Design Approach

After user testing and cross-functional stakeholder feedback, we ultimately settled on a more traditional approach that follows other best-in-class products and better supports accessibility standards. This decision was driven by our commitment to creating an inclusive experience that works for all users while maintaining the platform's powerful capabilities.

Final ML Use Cases Interface

Business Value

Streamlined ML Adoption

Integrated Data Prep eliminates complexity for first-time and citizen data scientists, enabling faster adoption of ML workflows and unlocking new use cases.

20% Reduction in data setup time
90% Platform retention rate

Accelerated Time-to-Value

By integrating data prep directly into DataRobot, it accelerates time-to-value and reduces onboarding friction.

25% Faster time to production
2x More active ML projects

Measurable Business Impact

By designing and delivering Integrated Data Prep, I helped streamline machine learning workflows, reduce operational complexity and costs, and enable broader adoption across teams. This work contributed to strengthening the product's position as a robust end-to-end AI solution, supporting key revenue and adoption goals.

Explore More Projects

Discover other innovative solutions in data management and user experience design.

View All Projects