Trusted Agents for
Retail and Consumer
Prometheux helps retail and consumer goods teams put AI to work across customers, merchandising, supply and stores, on your own data and business rules. Assistants, agents and teams get the same answer with the evidence behind it, and data is processed in place across your commerce, ERP, POS and warehouse systems: nothing to migrate.
| product | add-on | via |
|---|---|---|
| Ceiling fan F-210 | Wall control WC-3 | light kit LK-2 |
| Ceiling fan F-210 | Light kit LK-2 | direct |
| Fan F-118 | Downrod D-12 | direct |
Proven in production
Plenty of Data.
Too Few Answers.
Customer, product, stock, price and store data sits in different systems owned by different teams. AI pilots look good in a demo, then stall when the answer has to be right, consistent and explained.
Where Retail Teams
Start
Pick the area you own. Each use case runs on the same rules and on data left where it lives, so the second one reuses what the first one built.
Customers & Personalisation
Know each customer across channels and give them answers and offers your teams can stand behind.
Merchandising & Pricing
Product, price and promotion knowledge your merchandisers can read, applied the same way in every channel.
Supply & Inventory
Connect demand, stock, orders and suppliers so planners and the shelf see the same picture.
Stores & Operations
Give store and operations teams the same answers as head office, with the reason attached.
The Questions Other
Tools Can't Answer
Everything a Project Needs
What does a shopper need to install this ceiling fan, which parts fit this model, and which are in stock at their store? A recursive walk from project to product to compatible part. See recursion.
One Shopper, Many Records
Which online accounts, loyalty IDs and store receipts belong to the same shopper, across systems that never shared a key? Every match cites its rule. See entity resolution.
From Batch to Basket
If a supplier's batch is recalled, which SKUs contain it at any tier of the bill of materials, which stores and warehouses hold them, and which orders included them?
Supplier Exposure, Totalled
Which products depend on a supplier, directly or through its own suppliers, and how much revenue sits behind each? Recursion with aggregation totals it at every tier, with lineage to the source rows.
One Logic Layer
Across Commerce
Typical personalisation, planning and customer-data platforms each copy your data into their own store first. Prometheux sits between your catalog, customer, supply and store data and the teams, apps and agents that use it, and runs the rules where the data already lives.
Analytics
Merchandising, planning and store dashboards, graph and BI
Integrations
Back into the databases behind search, commerce, ERP and CRM
Agents
Shopping, service and operations agents, deterministic answers
Models
Define the ontology, read answers, and power analytics, integrations and agents
Your retail data
Catalog, inventory, orders, pricing, loyalty, store and supplier data, connected in place. Nothing to migrate.
-
Commerce, Supply and Stores
- Merchandising, customer, supply, pricing and store use cases on one set of rules
- Shopping, service and operations agents call deterministic rules over MCP and API, with the derivation behind every answer
- Results written back to the databases behind search, commerce, ERP and CRM, on a schedule or when data changes
-
The Retail Logic Engine
- One model of products, customers, stores, suppliers and prices across systems, instead of a copy per tool
- Recursion with aggregation and graph algorithms: compatibility chains, bundles, multi-tier supplier exposure, store and fulfilment networks; alongside vector search, not instead of it
- Same input, same output, with lineage to the rules and source rows
-
Cloud Agnostic
- Run in your own cloud next to the catalog, inside Snowflake and Databricks, or on your own Kubernetes
- Scale horizontally on clusters as catalogs, order history and store data grow, with optional GPU acceleration
- 1-click deployment anywhere: EKS, AKS, GKE, Databricks Clusters, Snowpark Containers
-
Retail-Data Agnostic
- Read and write catalog, inventory, order, pricing, store and customer data in any database, format and schema
- Connect warehouses, operational databases, object storage, vector stores and commerce APIs across clouds and on-premises
- No new customer-data store to migrate into and keep in sync
Same Layers.
Better Built.
Map a typical retail stack, a personalisation platform here and a planning tool there, to Prometheux and the difference is where the work happens: in place, in rules you own, with an answer you can check.
Exact Answers.
Not Lookalikes.
A Fortune 100 retailer's AI shopping assistant needed exact, explainable product answers across millions of products and thousands of attributes. Vector search alone kept returning items that sounded right but didn't fit. Prometheux connected the catalog in place, inside the retailer's own cloud, and added rules over the real relationships between products, parts and projects. On real customer questions it retrieved better answers than vector search alone, with each product fact traced to its source.
Owned, Not Vendor-Bound
Product, customer and supply logic stays with you, in rules your merchandisers and engineers can read and version, not inside a vendor's model.
Deterministic and Explained
The same question gives the same products in every channel, with the rules and source rows behind each one.
Built for Catalog Scale
Rules run as distributed, parallel jobs, with optional GPU acceleration, on compute you already have. See high-performance engine.
Portable by Design
Run in Prometheux cloud, your cloud, your own Kubernetes, or inside Snowflake and Databricks. Your customer data stays where it is.
Codify product knowledge once. Reuse it in search, service and planning. Give agents answers they can explain.
One Use Case.
Then the Next.
Connect in Place
Point Prometheux at the data for one use case: catalog, inventory, orders, pricing or store data. The data is processed where it lives; nothing is copied.
Codify the Knowledge
Your merchandisers, planners and pricing teams, our forward deployed engineers and the AI assistant turn compatibility, reorder, pricing and promotion policies into readable rules in days, not months.
Run One Use Case
Product discovery, supplier exposure, promotion eligibility or store range checks, on live data, for teams, assistants and agents alike.
Expand Across the Business
Each new use case reuses the rules already written, from catalog to customers, supply, pricing and stores, without rebuilding context for every tool.
Questions,
Answered
Do we have to replace our search, personalisation or planning tools?
No. Keep them. Prometheux works alongside vector search, planning and pricing engines: they keep doing what they do, rules decide what fits, what is eligible and why, and results are written back to the systems your teams already use.
Is this only for e-commerce and recommendations?
No. The same rules serve merchandising and catalog, customer and loyalty, supply and inventory, pricing and promotions, and store operations. Most retailers start with one use case and reuse the rules for the next.
Is this another customer-data platform?
No. Prometheux doesn't build a new profile store. It reads catalog, customer and order data where it already lives, across clouds and on-premises, and runs the rules there. There's nothing new to migrate or keep in sync.
Does our customer data leave our cloud?
It doesn't have to. Prometheux can run in your own cloud, inside Snowflake and Databricks, or on your own Kubernetes, under your controls. We're SOC 2 Type II.
Who keeps the rules up to date as the catalog, stock and prices change?
The rules run on live data, so new SKUs, stock levels and prices are picked up on the next run, on a schedule or when the data changes. When the logic itself changes, the AI assistant drafts the update in plain language and your team reviews it before it runs.
What does it cost at our catalog size?
It depends on your data and use cases, so we measure it on yours rather than quote a generic figure. Most retailers start with one use case, prove the value, then expand.
Related Capabilities
Bring One Retail
Use Case.
Customer 360, pricing rules, supplier exposure or store ranges, on your own data. Our forward deployed engineers will run it with you, in place, on your infrastructure.
