Kickstarting a Leading Chemical Firm’s AI Transformation

Requirements Analysis

Preceding technical implementation, our [at] experts conducted a thorough requirements analysis workshop. For this purpose, we gathered detailed insights into working practices, needs, and pain points across the entire AI development and operations lifecycle through interviews with data scientists and based on citizen data scientists’ needs, i.e., chemists, buyers and other business users who use AI models in daily work.

Technical Implementation

These interviews resulted in more than 60 clear user stories that served as central foundation for platform architecture. Building on this foundation, our experts carried out structured technology evaluations and formulated concrete recommendations for the data science teams’ future working environment, spanning both backend and frontend capabilities. Selected options were then validated through rapid prototyping to reduce decision risk and confirm technical feasibility early. Based on that, a future-ready MLOps platform was designed, aligned with tried-and-tested best-practices.

Roadmap

Finally, a concrete migration roadmap was defined for the transition from the current fragmented setup towards the new platform. This roadmap addressed not only technological aspects but also required changes in working culture, governance, and operating model. This rendered a clear and realistically executable path for the firm to embed the new platform sustainably within the organization, and operate it at scale.