Demand
Managing scale and complexity in a high-volume B2B environment
Imes Dexis is a leading technical wholesaler managing a catalogue of over 300,000 products, serving multiple industries through a dense operational model combining back-office teams, product experts and geographically distributed branches.
As volumes grew and communication channels multiplied (EDI, web services, email, fax), operational pressure increased. Teams had to constantly switch between systems, information was fragmented, and knowledge sharing across locations became increasingly difficult.
The challenge was clear: how to maintain speed, accuracy and margin control at scale - without adding operational burden or headcount.
Imes Dexis needed a way to:
- streamline order intake and communication flows
- preserve and leverage expert knowledge across the organization
- improve pricing decisions in a complex discount and segmentation landscape
- prepare its ERP and processes for long-term automation and intelligence
Design
Building an AI-driven foundation for operational intelligence
we+ partnered with Imes Dexis to design an AI/ML roadmap focused on business processes first, not technology for its own sake.
The initial design focused on creating a unified view of communications and processes:
- all incoming and outgoing messages were analysed and linked to specific business processes (order intake, product information, customer segmentation)
- optimization goals and KPIs were defined upfront to measure impact
- Apache Kafka for real-time data streaming and analysis
- custom language models, complemented by Gemini for generative AI use cases
- semantic links between communication data, product information and customer behaviour
From a technical perspective, the architecture combined:
This foundation allowed Imes Dexis to move from isolated automation to process-level intelligence, enabling smarter decisions across operations, pricing and knowledge management.
Deliver
From AI pilots to tangible operational results
The AI/ML initiatives were progressively embedded into daily operations:
- Order intake automation
Communication classification and process linking reduced manual intervention and significantly accelerated order processing.
- Knowledge sharing at scale
Automatic tagging, classification and distribution of information enabled experts and local teams to access the right knowledge at the right time.
- Pricing intelligence
Machine learning models analysed historical invoiced order lines and customer segments to recommend safe price ranges - protecting margins while reflecting customer sensitivity. Pricing decisions shifted from static lists to data-driven distributions.
- Product information & compliance automation
AI automated product onboarding, categorization, FAQ generation, SEO content and semantic search.
Safety sheet checks moved from spot checks to 100% automated control, ensuring compliance without manual effort.
- Broadcasting & communication
Product updates and safety information were automatically shared with relevant internal and external stakeholders.
Drive
Measurable impact today, scalable foundations for tomorrow
The results delivered immediate and long-term value:
- 50% reduction in order intake time, improving speed and reliability
- 40% increase in usable knowledge during the offer phase
- Operational scalability achieved without additional personnel
- Reduced personnel costs through automation of safety sheet management
- Stronger margin control through dynamic, segment-based pricing intelligence
Beyond these gains, Imes Dexis established a future-ready foundation. Planned next steps include AI-driven ERP optimisation for lead-time estimation and restocking, combining historical data, real-time inventory and predictive models.
With this approach, Imes Dexis transformed AI/ML into a structural accelerator of performance, enabling growth, resilience and continuous optimisation in a complex B2B environment.