Platform architecture

How GetFieldMind AI works.

GFM is built in layers. Each layer depends on the data organized by the layer below it, forming a coherent, purpose-built structure for field-service operations — first applied to pool service maintenance.

The five layers

At the foundation of everything sits Ingestion, which captures and organizes the raw field data generated by technician visits. Every other module consumes, directly or indirectly, the data Ingestion organizes — which is why it comes first.

Growth: Modules focused on business expansion — lead generation, institutional content, and management reporting.
ReportsLeadFinderContentPulse
Operation: Automates operational management processes, such as service dispatch and maintenance reminders.
DispatchCRMFlowGuard
Intelligence: A query and assistance layer — lets managers and technicians ask questions in natural language and get answers and guidance.
InsightPoolMind
Field Monitoring: Tracks operational indicators in real time (chemicals, water level, inventory, parts, technician productivity) and triggers alerts.
PulseCheckChemAlertGreenEyeLevelWatchStockSensePartsWatch
Base Layer: Receives and organizes raw field data and applies AI to evaluate and cross-reference the information consumed by every other module. The foundation for all other modules.
Ingestion
About the flow: a dashed line below the base layer, in the original architecture diagram, represents tools that client companies already use (e.g., Skimmer, other management software), which feed Ingestion with external data. LeadFinder feeds new contacts/leads back into Ingestion, closing the loop on the platform's database.

Infrastructure architecture

Complementing the functional architecture of the modules, this section describes the technical infrastructure that supports the platform at a conceptual level (without naming specific tools or vendors, for business confidentiality).

Server01Automation core
ServerMsg1 / ServerMsg2Messaging channels (redundant)
ServerMgmManagement & monitoring
Central DatabaseShared across all components
The platform runs on a main server responsible for the automation core, with dedicated servers for messaging channels (redundant with each other) and a management/monitoring server. All components connect over the internet and share a central database. This redundancy ensures communication with customers and technicians stays operational even if one messaging channel fails. Continuous monitoring of this infrastructure is described in detail in the InfraGuard module.

The 15 modules

Each module below links to a dedicated page describing its objective, how it works, who gets notified, and a real example from production.

Production

1. Ingestion

Captures the service-report emails sent by the field data-collection tool (Skimmer) after every visit and organizes them into a structured database, cross-referencing the information against a pre-registered CMDB (customers, technicians, pools, and products).

Production

2. PulseCheck

Tracks each technician's progress throughout the day in real time — how many pools have been serviced against the total scheduled — and alerts the manager about delays or pace deviations.

Production

3. ChemAlert

Checks the parameters recorded at each visit (chemical readings and filter pressure) and alerts the manager whenever a value falls outside the ideal range, highlighting the critical reading together with the chemicals applied during the visit.

Production

4. GreenEye

Analyzes the pool photos submitted by the technician during the visit and automatically identifies, through computer vision, when the water is green or shows signs of algae — alerting the manager immediately for action, along with a detection confidence score.

Production

5. LevelWatch

Monitors the pool water level recorded during the visit and, when it is detected below ideal, alerts both the manager (for internal follow-up) and the customer (with guidance on how to fix it) — preventing equipment damage from dry running.

Production

6. StockSense

Tracks chemical purchases from the SCP and PWP suppliers by automatically processing the purchase-confirmation email (invoice) to update inventory and consumption history — with no manual data entry required.

Production

7. PartsWatch

Consolidates every equipment issue reported by technicians (through the Skimmer checklist) that is still open, organized by customer and technician, giving the manager a complete view of pending repairs.

Production

8. Insight

Lets managers ask natural-language questions over WhatsApp about the platform's operational data (pools, readings, technicians, and more) and receive an automatically processed answer, without needing to consult reports or spreadsheets manually.

In Development

9. PoolMind

Provides a WhatsApp assistant that combines general pool-chemistry knowledge with customer-specific information (access details, particularities, history) to answer technician questions in the field, right when they're needed.

In Development

10. DispatchCRM

Centralizes service-call management and routing, letting managers create, assign, and track work orders (e.

Production

11. FlowGuard

Identifies which customers have an overdue filter backwash (30+ days) and sends a reminder directly to the responsible technician, prioritizing the most urgent cases on the day's route.

Future Roadmap

12. LeadFinder

Identifies homes with pools in a target region by cross-referencing public county data (API) with satellite image analysis, generating a list of qualified leads for 9Pool and GFM sales prospecting.

Future Roadmap

13. ContentPulse

Periodically publishes content about GFM across multiple channels — company website, LinkedIn, Pool Deck, and technical/academic outlets (Google Scholar, IEEE) — reinforcing market presence and formally documenting the platform's authorship and technical evolution.

Future Roadmap

14. Reports

Consolidates data from every module into periodic management reports — a weekly summary (every Friday) and a more complete monthly report — with statistics, charts, and key events, sent automatically to the manager.

Production

15. InfraGuard

Continuously validates the health of the platform's technical components (services, communication ports, messaging connection, database, and backups), giving the technical operator a clear confirmation that the environment is fully operational.

Interested in becoming a pilot partner?

GetFieldMind AI is currently expanding through selected pilot programs with service companies interested in improving operational efficiency and service quality through intelligent automation.

Contact GetFieldMind AI