How a Custom Fleet Dashboard Gets Built in 2 Weeks

The Problem: Three Systems, One Headache

A mid-sized logistics company in Northwest Arkansas was running on chaos.

Their fleet of 40+ vehicles was tracked across spreadsheets, legacy dispatch software that hadn't been updated since 2015, and phone calls. Lots of phone calls. Dispatchers spent 3+ hours per week manually entering data, cross-referencing systems, and trying to keep a real-time picture of where vehicles actually were. Drivers were calling in with status updates. Customers were calling asking where their shipment was. Management had no visibility into efficiency or idle time.

They'd looked at commercial fleet management platforms. $500/month per vehicle. Long contracts. Feature bloat they didn't need. And they'd still be locked into someone else's workflow.

They needed something simple: a custom dashboard that pulled data from their existing dispatch system, showed real-time vehicle locations on a map, alerted them to issues, and actually saved time instead of creating more work.

Day 1: Strategy Session

The strategy call is where we found out what actually mattered.

Not "build me an app." Real constraints: their dispatch database schema (we'd have to work with it), drivers using old Android phones (responsive design required), offline requirements (updates cached locally), and integration with their accounting system (each trip needed to log distance for reimbursement).

In 90 minutes we mapped:

Core entity model (Vehicles, Trips, Events, Alerts)
Data synchronization strategy (pull dispatch data hourly, real-time alerts for breakdowns/delays)
MVP feature set (map + list view, basic analytics, email alerts)
Phase 2 opportunities (driver mobile app, predictive maintenance, customer portal)

Scope: locked. Timeline: 14 days. Resources committed. No scope creep, no "while we're building it" feature requests. This was the deal.

Week 1: Foundation

We built the plumbing first — invisible to the end user but everything depends on it.

Day 2-3

Database Layer + API

Designed normalized schema for vehicles, trips, events. Built API endpoints for reading dispatch data, ingesting GPS updates, triggering alerts. Used event sourcing for the audit trail — every vehicle state change becomes an immutable event record.

Day 4-5

Data Sync + Ingestion

Built worker that pulls from their legacy dispatch system every hour, maps fields to our schema, detects deltas (vehicle status changed, trip assigned, etc), fires appropriate events. Handles disconnects gracefully.

Day 6-7

Core Dashboard UI

Built the main interface: map view with live vehicle positions, data table with trip summaries, filter/search. Used Mapbox for real-time tracking (updates every 30 seconds if GPS data available). Dark theme because they liked the mockups that way.

By end of Week 1: Dispatchers could log in, see all vehicles on a map in real-time, click any vehicle to see its trip history. No real data yet — just plumbing and UI.

Week 2: Features + Polish

Day 8-9

Alert System

Defined alert rules: vehicle stationary for 30+ mins outside a delivery zone (possible breakdown), trip delayed beyond expected completion, low fuel (if available from API), driver offline for 2+ hours. Real-time in-app notifications + email to dispatcher + SMS if critical.

Day 10-11

Analytics + Reporting

Added weekly summary view: total miles, trips completed, average idle time per vehicle, efficiency score (miles per hour active). Showed trends week-over-week. Data exports to CSV for accounting.

Day 12-13

Mobile Responsive + Testing

Optimized for tablets and old Android devices. Tested with their actual fleet — GPS devices, network conditions, offline scenarios. Fixed edge cases (what if a vehicle goes offline mid-trip, then comes back online?). Trained dispatchers on the interface.

Day 14

Deployment + Handoff

Migrated 6 months of historical data. Set up backup procedures. Deployed to their private server. Went live. Documented everything.

The Result: 3 Hours Back Per Week

First week post-launch:

Time saved: Dispatchers estimate 3-4 hours of admin work eliminated per week. No more manual data entry between systems.
Visibility: Management can now see real-time fleet status instead of calling for updates. Built predictive accuracy into routing decisions.
Efficiency: Identified that one vehicle was sitting idle 12+ hours per week (dispatch workflow issue, not the vehicle). Fixed the workflow, saved $2,400/month in vehicle costs.
Driver experience: Clear trip assignments + navigation integration meant drivers spent less time on phone confirming details.
Customers: Dispatch team can now give accurate ETAs to customers instead of "sometime this afternoon."

                    What matters: They're not running custom code just to say they have an app. They're saving real time, making better decisions, and eliminating manual work. The ROI will pay for 2 years of maintenance in the first 3 months.
                

Why This Took Exactly 2 Weeks (Not 6 Months)

Scope was non-negotiable. We didn't say yes to feature requests mid-project. Phase 2 exists for things we identified but deferred.
Architecture decision upfront. Event sourcing meant we could audit every state change without adding complexity to the core system. It took one day to explain and implement, saved two weeks of edge case debugging.
Real data sync, not fake integrations. We integrated with their actual legacy system immediately — no "mock API" followed by real integration later.
No new frameworks for the sake of it. Used boring tech that we knew inside and out. Node backend, React frontend, PostGIS for geospatial queries. Done.
Testing on actual hardware. Tested with actual Android devices and network conditions, not just a dev environment. Caught real bugs early.
One week buffer built in. We planned for 10 days of work + 4 days of buffer. The buffer turned into testing + optimization, not panic.

This Is What Matters to You

If you're running your logistics operation on spreadsheets and legacy software, you already know the cost: time, money, visibility, accuracy. Custom software built for your specific workflow doesn't have to take 6 months and $300k.

The constraints that made this project move fast are the same constraints that should apply to any custom app:

Clear scope defined upfront
Real integration with your existing systems
MVP first, phase 2 later
Testing on your actual equipment and network
Transparent timeline and budget

14 days. One focused team. One problem solved completely.

Got a Similar Challenge?

Fleet management, supply chain tracking, field operations software, delivery optimization — if you're juggling multiple systems and losing time to manual work, we can help.

Let's talk about your workflow in a strategy session. No pitch. Just honest conversation about what you're dealing with and whether custom software actually makes sense for your operation.

Book a Strategy Session