Live telemetry
Position, movement, distance, fuel, and device health become current fleet state instead of disconnected tracker messages.
Back to Work
Compass Solutions
GPS fleet-tracking SaaS with AI-augmented operations
Production fleet-management and GPS-tracking platform that helps businesses monitor vehicles, manage operational data, and act on real-time location-based workflows. AI-augmented modules surface insights from telemetry; the Compass Assistant lets operators ask fleet questions in plain language and get back structured answers with deep-links into the relevant pages.
Next.jsHonoTypeScriptPostgreSQLReal-time map (Mapbox)WebSocketsRBACAI Assistant (LLM)Tailwind CSSMulti-tenant architecture
10+
production modules owned
Real-time
GPS + telemetry tracking
Tool-calling
fleet operations assistant
Multi-tenant
fleet & organization scoping
Overview
Fleet operators don't need more dashboards. They need answers.
Compass is a GPS fleet-tracking platform built for businesses that need to monitor vehicles in real time, manage operational data, and act on location-based workflows — dispatching, geofencing, fuel monitoring, driver behavior, and predictive maintenance.
What sets it apart from a generic fleet-tracking SaaS is the AI layer. Compass ships AI-augmented modules for drivers, fuel, and maintenance that surface insights from raw telemetry without engineers having to query Postgres directly. The Compass Assistant takes that further: operators ask plain French/English questions like “which vehicles drove the most yesterday?” and get a structured answer with deep-links — no SQL required.
I work across the full stack on this product — web app, backend API, PostgreSQL data layer, generated API contracts, deployment workflows — and own the integration layer between real-time location data and the operational dashboards.
Operational model
From moving vehicle to accountable decision.
Compass connects two product surfaces—the fleet operator workspace and the partner portal—through one operational model for live state, incidents, access, and grounded answers.
Operational rules
Speed, geofence, braking, idle, maintenance, and fuel signals become evidence-backed alerts that teams can triage and assign.
Partner control
Client organisations, devices, SIMs, billing state, and provisioning stay manageable from a dedicated reseller-side workspace.
Grounded assistant
Natural-language questions resolve to fleet-specific tools and live records, returning useful answers with paths back into the product.
Operator · Overview
The whole fleet on one screen
Fleet status, distance, alert volume and fuel spend for the selected range, over a live satellite map of every active vehicle — then weekly activity, top distances, recent trips and the latest alerts underneath.
Operator · Vehicles
Every vehicle, and whether it is actually reporting
Online / offline / needs-attention split across the fleet, then a card per vehicle with make, plate, device ID, driver assignment and last-seen time. Grouping, sorting and search on top for larger books.
Operator · Alerts
Every incident, triaged and assignable
Speed, geofence, harsh braking and idle events, split by severity with mean response time and treatment rate. Each alert opens evidence — the measured value against the rule, the map position, the source sensor, coordinates — and a one-click intervene action.
Operator · Compass Assistant
Ask the fleet a question in plain language
LLM tool-calling agent scoped to fleet operations. It resolves plain French or English questions into the right tool calls — vehicles, alerts, trips, drivers, maintenance, fuel — runs them against the live database, and answers from tool output rather than from the model.
Partner · Partner fleet
879 vehicles across 666 client organisations
The reseller-side portal. Every device across every client organisation in one book: online, offline and never-reported counts, financing state for clients awaiting billing, and search across client, phone, plate, IMEI or SIM.
Partner · Device & SIM
Provisioning down to the SIM and the IMEI
Per-device record tying plate, IMEI, tracker model, SIM and carrier to the owning client and driver, with live position, odometer, document expiry, and an SMS command channel to the tracker itself.
Engineering Challenges
The hard parts I worked on
Real-time at fleet scale without melting the database
Solution · Telemetry arrives far more often than the product needs to redraw every screen. The system keeps current map reads and historical reporting on different paths, with buffered writes and aggregated records where the workload calls for them.
Outcome · Operators can move between current vehicle state and historical activity without treating them as the same query problem.
Multi-tenant access across vehicles, drivers, organizations
Solution · Users, organisations, vehicles, drivers, geofences, and reports all carry explicit ownership context. Query paths and team-management actions respect that scope rather than relying on the interface to filter records after loading them.
Outcome · The same organisation boundary is applied across operator screens, partner tooling, and reporting paths.
AI assistant that actually answers fleet questions
Solution · The assistant uses a fleet-specific tool layer for vehicles, alerts, trips, drivers, maintenance, and fuel. It answers from tool results and links operators back to the relevant product surface instead of presenting free-form model output as operational truth.
Outcome · Common questions can begin in natural language while the underlying data remains traceable to the product.
Engineering-friction reduction across the codebase
Solution · Introduced reusable interface patterns, structured service logic, migrations, seed scripts, type-checking, and build-quality improvements in the areas I worked on.
Outcome · New work can follow established form, access-control, and notification patterns instead of inventing another local convention.
Platform architecture
Fast live state without sacrificing durable history.
The system separates high-cadence location updates from historical reporting and keeps organisation scope explicit at every boundary—from the map to the assistant.
Next.js + Hono product
Next.js powers the operator experience while Hono provides the typed API layer behind maps, vehicles, alerts, reports, forms, and partner tooling.
Real-time telemetry path
Buffered updates and live channels keep the map responsive while protecting durable storage from high-frequency write contention.
PostgreSQL operations
Current state, historical records, and aggregated reporting remain queryable through tenant-aware service boundaries.
Tool-grounded AI
The assistant can access only registered fleet tools and scoped records, making its answers traceable to production data.
What I Deliver
Day-to-day at Compass
- 01
Built and refined dashboard, vehicle, driver, map, and fleet-management flows for clearer day-to-day tracking and improved operational visibility.
- 02
Worked across the full stack — web app, backend API, PostgreSQL data layer, generated API contracts, and deployment workflows — to strengthen end-to-end product delivery.
- 03
Reduced engineering friction by introducing reusable UI patterns, structured service logic, database migrations, seed scripts, type-checking, and build-quality improvements.
- 04
Owned the integration between real-time location data and operational dashboards so fleet managers can act on driver and vehicle state without leaving the app.
- 05
Shipped AI-augmented modules (Compass Assistant, AI driver/fuel/maintenance insights) that translate raw telemetry into operator-facing answers.
Building real-time SaaS for your team?
I work end-to-end on production systems with real-time data, multi-tenant access, and AI-augmented operations.