All projects · NEX
NEX Prism
The AI proposes every price and explains why; the team decides.
- Client
- NEX
- Year
- 2025 – present
- Industry
- Automotive / Tyre distribution
- Role
- PM · Consulting · UX-UI · Development
- Platform
- Custom-built internal web tool
- Integration
- Microsoft Dynamics 365 Business Central
- Users
- Pricing team
- Catalogue
- Over 90,000 references
Price recommendation
Every proposal explains why: current and suggested price, range and confidence, the factors at play, the retail price trend and a 30-day simulation.
Summary
NEX distributes tyres from a catalogue of more than 90,000 references. Prism is its pricing tool: every morning an AI engine recommends a price for each reference based on stock, rotation, seasonality and target margin, and the pricing team reviews, applies, adjusts or discards each proposal.
Screens
Data shown is anonymized.
Home
What the pricing engine did each day by family: how many recommendations it generated and how many were applied.
Catalogue
The B2B catalogue with cost, margin and the AI-suggested price on the same row, to apply several suggestions at once.
Pricing rules
The business rules the AI cannot skip: minimum margins, rounding, clearances and changes that need approval.
Activity
Who applied, adjusted or discarded each change, alongside the engine’s daily runs.
The problem
Keeping tens of thousands of prices current, across two channels and with very different margins per family, cannot be done by hand or left to an opaque algorithm.
- The catalogue is too large to review price by price.
- Stock, rotation and season change every week and prices have to follow.
- Any automatic change must respect minimum margins and be explainable.
Approach
The AI proposes and people decide. Every recommendation comes with its reasoning, range and confidence level, and goes through business rules the engine cannot override. Large changes need approval and everything is logged.
What we built
Explained recommendation
Current and suggested price, recommended range and confidence, the factors behind the proposal, the retail price trend and a 30-day simulation of units, margin and stock cover.
Daily engine by family
Every morning the engine generates recommendations for B2B and B2C, and the home screen shows by family how many were generated and how many applied.
Catalogue with inline suggestions
Cost, margin and suggested price on the same row, with filters for low margin, high stock or slow rotation, and several suggestions applied at once.
Rules the AI respects
Minimum margins per family, commercial rounding, end-of-season clearances, variation limits and manual approval for large changes.
Activity and traceability
Who applied, adjusted or discarded each change and why, next to every engine run.
Stack
- .NET
- React
- MariaDB
- Business Central
Visual system
A working white and a navy blue that marks everything the AI proposes.
- Prism navy#0D2063
- White#FFFFFF
- Surface#F6F7F9
- Border#E8E9EE
- Up#16A34A
- Down#DC2626
- 318,00 €
Inter Tight
Interface and large price figures