PON: Moving Your Data Science Stack to Dataiku

PON has successfully migrated its data science stack to Dataiku in order to manage the data of seven robust business activities (i.e., automotive, bikes, equipment and power systems, industrial mobility, agriculture products and services, Ponooc, and Move Amsterdam) to work more effectively. The most impressive part? They were able to transform and optimize operations to run on the platform within all seven major activity areas in just three weeks.
Watch Video

A Fast but Thorough Data Platform Migration 

Starting as a modest family business over 125 years ago, PON has come a long way but, through many changes and lots of growth, they have remained true to their mission of moving people to a better world. The global conglomerate based in the Netherlands offers mobility products, services, and solutions across seven different major activity categories and, in order to manage and deliver these robust offerings, employs approximately 16,500 individuals in 65 different countries. As it turns out, moving people to a better world is no small feat and the data accumulated along the way through these processes is, as you may have guessed, exponentially vast. 

The 7 Activity Areas Migrated in 3 Weeks

  • Automotive
  • Bikes 
  • Equipment and Power Systems
  • Industrial Mobility 
  • Agriculture Products and Services 
  • Ponooc (A Venture Capital Fund) 
  • Move Amsterdam (Experience and Exhibition) 

PON’s previous, restrictive code-only platform did not offer the foundation or flexibility that these expansive offerings and their heavy data influx required. Therefore, a migration to the more accessible, collaborative, and powerful environment of Dataiku was a natural and essential next step to support the organization as it continues to scale.

Some of the use cases across these activities that are now managed with Dataiku include but are not limited to client churn prediction analysis, stock management, advanced product analytics, warranty claim detection, website personalization, and customer lifecycle analysis. 

PON’s Precursory Steps for Successful Migration 

When asked about the keys to their successful transformation, PON points out three fundamental factors:

  1. Not remaining in an exclusive data team huddle and reaching out to different teams (more on this below) 
  2. Identifying high-value projects and choosing a platform that caters specifically to top-priority business objectives and needs across the organization 
  3. Making decisions by trying to predict potential pitfalls that will need to be addressed

Before migrating they asked around about the varying needs of individuals across the organization beyond the data science team: business unit directors, cloud specialists, privacy/security officers, field engineers, etc. These various stakeholders made lists of must-haves and want-to-haves that they desired in a platform. And after determining which of their requests served the long-term visions of the organization, PON was able to choose the platform that best fit those varying needs and skills — Dataiku. After choosing the right platform, the next step was to figure out where on the maturity curve PON resided in order to know exactly what training and changes would be needed. Dataiku offers resources and support that help organizations identify their maturity level. 

Following the maturity assessment was the process of deciding where to direct focus by listing out each offering and scoring the products and services by priority and estimated impact. PON also took special care to pinpoint high-risk areas. By doing this, they allowed themselves the ability to evaluate and adjust migration stages, proactively addressing possible pitfalls.

PON efficiently and effectively transformed their data processes by setting aside time to dedicate to these crucial steps. 

Not Stopping There — Learning and Growth Continues 

PON hasn’t stopped moving forward after migrating. Teams across the organization are involved in continued training programs on new Dataiku functions, workshops on MLOps, and collaborative working sessions with Dataiku experts to discover, understand, and apply advanced analytics tools. Through this sustained learning, the organization is able to address business problems across all seven growing business areas with the latest available solutions. 

A Traditional Organization + a Modern Data Science Practice: The Oshkosh Story

Learn about the culture of data science at scale that Oshkosh Corporation has developed using modern software tools like Dataiku — a culture that has enabled significant cost savings and performance improvements over a variety of solutions for the business and their customers.

Read more
Watch video

Fives Group: Driving a Client-Focused Digital Transformation

Fives Group boosts efficiency and drives innovation by integrating AI throughout their internal processes as well as within externally facing client-focused processes.

Learn More

Improving Manufacturing
Processes with Essilor

See how one manufacturing company, Essilor, uses Dataiku to harness large, heterogeneous datasets and develop a robust predictive maintenance solution.

Learn More
Watch video

SLB: Powering AI and Analytics Across the Energy
Value Chain

By empowering their employees and customers to use data for everyday decision making, SLB is reshaping the future of energy.

Learn More
Watch video

Etihad Airways: Driving Business Agility with Data

Etihad Airways leveraged data to face the pandemic and stay true to its positioning of being customer centric. Hear from the data science and AI team about how data played a pivotal role in shaping tomorrow's strategy for the leading airline.

Learn More