Nexus: Intelligent Nerve Center of Connected Retail
Nexus is a multi-tenant, enterprise-grade AI analytics, data visualization, and system management platform for retailers. It unifies data across thousands of stores and millions of shoppers—providing deep visibility into shopper demographics, product affinity, value segmentation, loss prevention activity, and in-store traffic patterns.
Accessible from desktop or mobile, Nexus empowers executives, managers, and IT teams to build descriptive, predictive, and prescriptive analytics models from both structured and unstructured data, driving faster, data-driven decisions. Powered by Generative Business Intelligence and Large Language Models, Nexus lets any business user interact with retail data using natural language—transforming static reporting into proactive, enterprise-wide intelligence.
Nexus Platform Architecture Overview
Platform Modules
Analytics & Intelligence
Nexus Dashboard
Multi-store intelligence, unified in one view. Gen BI and LLM-powered analytics that transforms complex retail data into actionable insights via natural language queries, providing instant visualizations, forecasts, and recommendations without technical expertise.
Configuration & Control
Nexus Admin Portal
Control hub for the AI retail ecosystem. Self-service platform for onboarding data, configuring store environments, managing users, and defining the operational parameters that guide MindGlyph’s AI applications across the enterprise.
Demographic Profile
Population-level insight at store-level precision.
Product Affinity Analytics
Behavioral intelligence behind every category.
Shopper Value Segmentation
RFM scoring for the modern retail footprint.
In-Store Traffic Patterns
Spatial intelligence for store performance.
Loss Prevention Metrics
Shrink intelligence across the enterprise.
Merchandising Analytics
Data-driven space planning and realograms.
Demographic Profile
Delivers granular demographic insights—age, gender, ethnicity—by aisle, category, or store, enhanced with hyperlocal census overlays and predictive queries to analyze current shopper composition and forecast future profiles, enabling optimized hyper-local assortment and pricing strategies.
- Granular aisle-level demographics: Age, gender, and ethnicity data segmented at the aisle, category, and store level for precise assortment decisions.
- Hyperlocal census overlays: Shopper composition data enriched with external census data to contextualize in-store patterns against surrounding market demographics.
- Predictive profile queries: Forecast how shopper demographics will shift over time to enable proactive assortment and pricing strategy adjustments.
- Hyper-local assortment optimization: Demographic insights feed directly into planogram and assortment decisions, ensuring each store reflects its actual shopper base.
Product Affinity Analytics
Combines dwell time, behavioral signals, purchase history, and RFM data to uncover true product affinity, enabling AI-driven recommendation engines to deliver personalized offers, coupons, and promotions.
- Multi-signal affinity modeling: Fuses dwell time, path traversal, purchase history, and RFM scores into a composite affinity index per shopper and product category.
- AI-driven personalization: Recommendation engines use affinity data to surface relevant offers, coupons, and product suggestions in real time.
- Cross-category discovery: Latent shopper interest in unpurchased products—measurable through movement patterns, dwell time, and visit frequency—reveals a deeper layer of affinity that purchase-based recommendation models cannot predict.
- Promotion targeting: Affinity scores underpin targeted promotion delivery through the Dealz! engagement platform.
Shopper Value Segmentation
Performs real-time segmentation of shoppers by recency, frequency, and monetary value across stores and product categories to identify high-value segments, track evolving behavioral trends, and deliver personalized promotions to top shoppers while re-engaging fading ones.
- Real-time RFM scoring: Shoppers are continuously scored across recency, frequency, and monetary dimensions at both the store and category level.
- High-value segment identification: Surfaces the top shopper tiers by revenue contribution, enabling prioritized retention and reward strategies.
- Fading shopper re-engagement: Detects declining engagement patterns and triggers targeted re-engagement promotions before shoppers churn.
- Behavioral trend tracking: Monitors how segment compositions shift over time to detect emerging shopper patterns and adapt merchandising accordingly.
In-Store Traffic Patterns
Delivers interactive heatmaps overlaid on store floor plans that visualize shopper traffic, dwell times, and headcount, enriched with multi-store comparisons, demographic data, product affinity, and hyperlocal factors like weather and local events to drive smarter, data-driven decisions on layout, staffing, and merchandising through causal analysis.
- Floor plan heatmaps: Interactive traffic density maps overlaid directly on store floor plans, updated in real time to reveal hot zones and dead spots.
- Dwell time analysis: Measures shopper time-at-location by aisle and zone to quantify engagement intensity across the store.
- Multi-store comparison: Benchmarks traffic patterns across locations to surface performance outliers and identify best-practice layouts.
- Hyperlocal enrichment: Traffic data is augmented with external signals including weather patterns and local events for causal analysis.
- Staffing optimization: Traffic flow data informs labor scheduling and service positioning for peak periods.
Loss Prevention Metrics
Aggregates historical theft incidents, suspect profiles, frequently targeted merchandise, and multi-store patterns into detailed, federated statistics and searchable, video-linked reports, serving as a central hub that enables security teams to review and share imagery, biometric signatures, suspect descriptions, and video evidence across locations for multi-store triage and proactive loss prevention.
- Federated incident database: Centralized repository of theft incidents across all store locations, enabling enterprise-level shrink pattern analysis.
- Suspect profile management: Anonymized face embeddings, appearance signatures, and vehicle fingerprints are stored and searchable across the network for cross-location identification.
- Video-linked evidence reports: Each incident record links directly to associated video footage for rapid review and evidentiary packaging.
- Multi-store triage: Security teams can share imagery and suspect data across locations to coordinate proactive interception of repeat offenders.
- High-risk merchandise identification: Frequently targeted SKUs are surfaced for enhanced protection measures or placement adjustments.
Merchandising Analytics
Visualizes AI-generated planograms, detailed product-level realograms, and VR A/B testing outcomes to refine space utilization and assortment decisions across formats and regions.
- Planogram visualization: Renders Optima-generated planograms as interactive, product-level shelf diagrams within the dashboard for review and approval workflows.
- Realogram comparison: Side-by-side views of planned versus actual shelf conditions, surfaced from computer vision shelf audits.
- VR A/B test outcomes: VisoLab behavioral testing results are surfaced and compared within the Nexus analytics environment for unified decision-making.
- Cross-format analysis: Space utilization metrics are benchmarked across store formats and regions to surface performance gaps and best-practice layouts.
Data Access SDK
Secure, local access to enterprise data without moving it off-platform.
Merchandise Inventory
Universal catalog for every store’s SKUs and product attributes.
Merchandise Placement
Spatial blueprint for precise shelf intelligence mapping.
Device Management
The map behind Vision AI — camera-to-zone association.
Identity & Access Management
Enterprise governance for distributed retail teams.
Configuration Management
Integration layer for omnichannel connected retail.
Shopper Characteristics
Behavioral signals that power cognitive models.
Promotion Management
Campaign intelligence for dynamic retail.
Merchandising Optimization
AI-generated planograms for every store format.
Billing & Payment
Automated commerce for the platform economy.
Data Access SDK
Enables Nexus to connect directly to a retailer’s own data without moving it off-platform, supporting all major self-hosted and managed time-series, vector, and graph databases to power advanced analytics and AI-driven insights, while keeping all data under the retailer’s control and ensuring secure, high-performance integration with enterprise-scale analytics pipelines.
- Zero data movement: Analytics queries execute directly against the retailer’s infrastructure—data never leaves the retailer’s environment.
- Universal database support: Compatible with all major self-hosted and managed time-series, vector, and graph databases.
- Enterprise-scale pipeline integration: Designed for high-throughput production analytics workloads across thousands of stores.
- Data sovereignty: Retailers retain full control and ownership of all data processed by MindGlyph AI applications.
Merchandise Inventory
Maintains a consolidated repository of SKUs, categories, and product attributes to enable accurate product monitoring and store-level assortment intelligence across the enterprise.
- Centralized SKU repository: Single source of truth for all product identifiers, categories, and attributes across every store location.
- Assortment intelligence foundation: Product catalog data feeds into AI models for accurate planogram generation and compliance monitoring.
- Continuous catalog synchronization: Keeps product data current as new SKUs are introduced, discontinued, or reformulated.
Merchandise Placement
Defines aisle, rack, and shelf positions of items by store, ensuring AI model outputs can be precisely mapped to real merchandise positions for accurate planogram compliance and spatial intelligence.
- Store-level spatial mapping: Each product’s aisle, rack, and shelf coordinates are defined per store, creating a precise physical reference layer for AI models.
- Planogram-to-reality anchoring: Placement data ensures AI-generated planogram outputs correspond exactly to real shelf positions in each store.
- Compliance monitoring support: Spatial definitions enable computer vision models to detect deviations between planned and actual product placement.
Device Management
Associates each camera with store zones—entry, exit, aisle, self-checkout, or parking—to ensure the right AI models run on the right video streams, maximizing both detection accuracy and computational efficiency.
- Camera-to-zone mapping: Each physical camera is associated with a specific store zone type—entry, exit, aisle, self-checkout, or parking—defining its AI model assignment.
- Automated OTA updates: AI models and device firmware are updated over-the-air automatically, ensuring every camera across the enterprise runs the latest detection algorithms and security patches without manual intervention.
- Multi-store device inventory: Centralized management of all camera hardware across the enterprise with status monitoring and configuration tracking.
- Streamlined deployment: New store additions are onboarded through a guided device registration flow that maps hardware to zones without manual coding.
Identity & Access Management
Provides granular user roles, access privileges, and workflow permissions for secure and compliant system operation across distributed retail teams, regional offices, and third-party integrators.
- Granular role definitions: User roles are defined at the function level—executives, store managers, security analysts, IT admins—each with precisely scoped permissions.
- Workflow permission controls: Approvals, escalations, and data access gates are enforced through configurable workflow rules.
- Multi-tenant isolation: In enterprise deployments, data and access are fully isolated between organizational units, regions, and franchisees.
- Compliance support: Audit logs and access histories support regulatory compliance and internal governance requirements.
Configuration Management
Generates AR tracking markers for Dealz!, and defines cloud API endpoints for unifying in-store data with retailers’ backend systems and e-commerce sites, delivering a seamless omnichannel experience.
- AR marker generation: Automatically produces the tracking markers used by the Dealz! app for in-store AR product experiences and proximity promotions.
- API endpoint configuration: Defines the cloud API connections that route in-store behavioral data to retail backend systems including ERP, POS, and CRM platforms.
- E-commerce unification: Links in-store data streams with online channels to power unified customer profiles and omnichannel personalization.
- Self-service configuration: No-code configuration interface lets IT teams define integrations without custom development.
Shopper Characteristics
Sets behavioral thresholds for traffic flow, movement speed, dwell time, and value segmentation patterns that enrich customer profiles and calibrate the accuracy of analytics and AI models to each store’s unique shopper base.
- Behavioral threshold configuration: Admins define store-specific thresholds for dwell time, movement speed, and traffic flow that determine how behavioral signals are classified.
- Segmentation parameter tuning: RFM segmentation breakpoints and value tier definitions are configurable per store or banner to reflect actual shopper economics.
- Customer profile enrichment: Configured behavioral signals are applied continuously to enrich customer profiles as new shopper interactions occur.
Promotion Management
Delivers generic, personalized, and proximity-based coupons, product recommendations, and instant promotions with custom usage rules for distribution through the Dealz! app, bridging campaign strategy with real-time in-store execution.
- Three promotion tiers: Generic promotions for broad audiences, personalized offers driven by affinity and RFM data, and proximity-based triggers activated when shoppers enter specific store zones.
- Custom usage rules: Redemption limits, expiry dates, minimum purchase thresholds, and exclusivity rules are fully configurable per promotion.
- Dealz! distribution: All promotions are delivered through the Dealz! app, ensuring consistent shopper experience across the engagement platform.
- Campaign performance tracking: Redemption rates, shopper response, and incremental lift are reported back into the Nexus analytics dashboard.
Merchandising Optimization
Generates and refines AI-powered, hyper-local planograms using in-store data on traffic, dwell times, demographics, and product affinity to create dynamic, sales-optimized assortments. Enables immersive, crowdsourced 2D and 360° VR A/B testing to validate layouts before physical deployment.
- Hyper-local planogram generation: AI models consume store-specific traffic, dwell, demographic, and affinity data to produce planograms optimized for each individual store.
- In-store data fusion: Real-time behavioral signals from Nexus analytics continuously refine planogram recommendations as conditions evolve.
- VR pre-validation: Planogram variants are deployed into VisoLab crowdsourced VR environments for behavioral testing before physical execution.
- 2D and 360° testing modes: Supports both flat panoramic and fully immersive VR testing formats to accommodate any participant device.
Billing & Payment
Creates and manages invoices based on subscription tiers and usage metering, while handling one-time and recurring auto-payments through credit cards, bank transfers, digital wallets, and online payment platforms.
The Nexus Advantage
Nexus is the intelligence and control layer that unifies every data stream, every store, and every team—transforming fragmented retail operations into a single, continuously learning system that drives faster decisions, stronger performance, and deeper shopper understanding at enterprise scale.