Project · 2025-12-07 · 4 min read
Zenit Auto
See how Zenit Auto connects inventory data, purchasing recommendations, pricing review, traceable evidence, and an AI assistant for automotive-parts buyers.
Zenit Auto helps automotive-parts buyers decide what to order, what to discount, and where to investigate pricing or assortment gaps. It brings sales, inventory, and competitor data into one workspace, with AI-assisted recommendations, traceable calculations, and an assistant for exploring the reasoning behind each decision. Buyers review recommendations before preparing the final order.
The challenge
Automotive-parts purchasing involves thousands of products with different demand patterns, stock positions, and supplier constraints. Some products need replenishment, while others sit on shelves longer than expected. Competitor availability and pricing add another layer to each decision.
Buyers were combining sales, stock, and competitor reports by hand to understand what needed attention. Reviewing products individually took time, and a suggested action was difficult to assess without the information behind it.
The goal was to bring that information together and give buyers a practical way to move from identifying a problem to reviewing an action.
What we built
We built an inventory decision platform that connects product data, operational analysis, recommendations, and order preparation.
A shared inventory workspace
The workspace brings sales performance, inventory, and product information together. Buyers can inspect the position of a brand, explore its catalogue, and identify products that deserve closer attention.
Sales and stock history provide context for individual decisions. A low stock level means something different for a steady seller than for a product with occasional demand.
Purchasing recommendations
The platform uses demand estimates, available inventory, and purchasing rules to suggest replenishment quantities.
Buyers can review the planning target, the stock available to meet it, and the remaining quantity recommended for purchase. This makes the recommendation easier to assess than a quantity presented without explanation.
For example, a planning target of 80 units and 32 accessible units leaves 48 units to order. The evidence view lets the buyer inspect the inputs behind that calculation.
Pricing and assortment review
The platform also identifies situations where purchasing more stock may not be the right response.
Overstock recommendations support reviewing slow-moving inventory for potential markdowns. Competitor information helps buyers investigate price differences, availability gaps, and opportunities to expand their assortment.
These signals give the team a focused starting point for review while keeping commercial judgment with the buyer.
Evidence behind each decision
Recommendations include an explanation of the proposed action and the information used to produce it.
Buyers can inspect demand, inventory, decision rules, and available source records. Keeping these inputs connected to the recommendation helps the team question an unexpected result and understand what needs checking before proceeding.
An AI assistant in context
The assistant supports questions about products and recommendations within the workspace.
A buyer can ask why a quantity was suggested, explore the demand behind it, or request an explanation of the available evidence. This helps the team investigate decisions without repeatedly moving information into a separate conversation.
From review to an order proposal
Reviewed recommendations feed into a final-order workflow. Buyers can inspect the proposed products and quantities, then export the order information for the next step in their purchasing process.
The platform supports order preparation; a recommendation is not an automatically placed supplier order.
Our role
Perfectory designed and built the workflow connecting operational data with inventory analysis, AI assistance, and purchasing decisions.
Our work covered:
- Data ingestion and processing.
- Inventory and recommendation engines.
- Forecasting evaluation and backtesting.
- Product and recommendation interfaces.
- Evidence views and contextual AI assistance.
- Final-order preparation and export.
Results and value
Zenit Auto gives buyers one place to investigate inventory questions and review proposed actions.
Instead of starting every decision by combining separate reports, the team can work from a prioritized set of recommendations and inspect the supporting information where it is needed.
The result is a more structured purchasing workflow: identify what deserves attention, understand the reasoning, apply buyer judgment, and prepare the next action.
Get in touch