VSM — Virtual Store Manager

ALX is the store agent. It is Canary’s primary customer-facing differentiator.

What ALX holds

At any moment, ALX has access to:

• Real-time inventory and SKU locations (from T module ingestion of Counterpoint)
• Live promotions and margin targets (P module)
• Active LP alerts and exception queue (Q module)
• Staff availability by zone (L module, where active)
• Device status and sensor data across the store (A module)
• Demand signals from other store agents that day (back office hub)
• Store brand voice, policies, and service standards
• 130+ years of domain knowledge (for L&G: plant diagnostics, care protocols, seasonal calendars)

What ALX does

Customer-facing:

• Receives incoming customer Claude agent via MCP handshake
• Answers real-time inventory queries: “Is the 5-gallon Meyer lemon in stock?”
• Delivers domain-specific intelligence: diagnoses plant symptoms, recommends care protocols, suggests products
• Routes to human associates when the conversation requires it
• Settles transactions to customer accounts — touchless checkout
• Logs visit feedback and purchase context post-transaction

Operator-facing:

• Surfaces LP alerts to GM via Owl
• Reports exception queue to Fox for case management
• Feeds cross-store signals to the back office hub
• Monitors device heartbeats via Module A

The interaction model

Customer's Claude (phone / wearable)
         │
         │ MCP / HTTPS + SSE / JSON-RPC 2.0
         ▼
    ALX (store MCP server)
         │
         ├── check_inventory(item, store)
         ├── get_diagnosis(symptom, plant_type, region)
         ├── get_product_recommendation(diagnosis, store)
         ├── get_pricing(sku, customer_tier)
         ├── find_in_store(product, store)
         ├── authorise_transaction(customer_id, basket)
         └── log_visit(customer_id, outcome)
         │
         ▼
    NCR Counterpoint (via REST API / ODBC)

Two agents. One conversation. No human required unless ALX decides otherwise.

Why this is Canary’s differentiator

Every other analytics layer on Counterpoint is a dashboard that humans read. ALX is an agent that acts. The spine modules (T/Q/D/J) are the knowledge engine. ALX is the interface. A VAR that deploys Canary doesn’t just give retailers better reports — they give retailers a store agent that any customer’s AI can talk to.

No other NCR Counterpoint VAR has built this. No other entity will build it before the window closes.

Domain knowledge: the L&G example

For Lawn & Garden deployments, ALX encodes vertical domain knowledge that has never been in a machine-readable system:

Knowledge type	Example
Plant diagnostics	Symptom → diagnosis → treatment tree for 500+ species
Zone-aware care	Sunset Climate Zone filtering on all recommendations
Seasonal availability	Month-by-month planting calendars for SoCal zones
Soil and amendment protocols	By plant type + current soil conditions
Pest identification	Symptom → pest → organic/conventional treatment
Companion planting	Compatible species combinations

An Armstrong customer who walks in and says “My Meyer lemon leaves are yellowing” gets a 136-year-old horticulture firm’s answer in their pocket before they reach the soil aisle.

→ Architecture — the full stack → Verticals — L&G playbook → Armstrong Garden Centers — proof case