Logistics
every vehicle,live on one screen
Operations software where dispatch, the fleet, and the people waiting on them see the same status at the same time. Built for the moment something goes wrong.
Illustration: three delivery vehicles looping routes from a depot on a schematic map, marking each stop as delivered and counting down the time to their next stop.
- concurrent users on real-time systems
- 10K+
- critical-alert response time
- −65%
- concurrent live conversations
- 1,000+
- venues running fleet and scheduling ops
- 5
01 · The hard part
Where operations lose the thread
Logistics problems are rarely about the map. They are about who knew what, and when.
Dispatch works from stale status
Positions and job states update on a refresh, so decisions are made on where things were, not where they are.
Alerts arrive after the problem
A failed delivery or a device fault reaches the right person by phone call, long after it could have been fixed.
Schedules live in spreadsheets
Vehicles, drivers, and time slots are balanced by hand, and double bookings are found on the day.
Crews on patchy connections
Field apps that assume a perfect signal lose updates exactly where the work happens.
02 · What I build
What I have built for operations
Three pieces of operational software that ran in production, plus the one I would add next.
Fleet and scheduling operations
A back office that owns the fleet, the calendar, and the reporting, so the operator stops reconciling three tools.
Shipped in: Track Booking Platform- Fleet management
- Event and slot scheduling
- Operational reporting
- CRM alongside bookings
- Typed API clients
- Route guards and validation
Real-time status and alerts
Status changes pushed to every screen that cares, with emergency alerts routed to the right people.
Shipped in: Smart Living Dashboard- Socket.IO live updates
- Push notifications
- Emergency alert routing
- Audio and video calling
- Serverless backend on AWS SAM
- Sub-second message delivery
Device and sensor monitoring
Device APIs wired into mobile apps so faults raise an alert instead of waiting for someone to notice.
Related work: Mobile development service- Device API integration
- Mobile monitoring apps
- Critical-alert escalation
- Device status history
- Ionic cross-platform clients
- Push to the on-call phone
Route planning and ETAs
Not shipped yet. The approach: a maps provider for routing, geofenced stop events, and ETAs recalculated from live positions on the same real-time channel as status.
- Maps provider routing
- Geofenced arrivals
- Live ETA updates
- Customer tracking links
03 · Stack
The tools behind the work
Real-time transport first, then the services and apps that sit on it.
- Real-time
- Socket.IO
- WebSockets
- Firebase
- OneSignal
- Backend
- Node.js
- NestJS
- Django
- PostgreSQL
- Celery
- Mobile & devices
- Ionic
- Angular
- Google Nest SDM APIs
- Cloud
- AWS Lambda
- SNS
- SES
- AWS SAM
- CloudWatch
04 · Reliability & security
Keeps working when the signal does not
Operations software is judged on its worst day. These are built in from the start.
Tokens that refresh themselves
Long shifts do not end in a surprise sign-out; sessions refresh in the background.
Alerts with an owner
Every alert routes to a role, and escalates if nobody acknowledges it.
A history of every status change
Who changed what and when, kept with the job, so disputes are settled from the record.
Scoped access per role
Drivers, dispatch, and customers each see their slice, enforced at the API.
Monitoring from the first deploy
Queues, sockets, and API latency are on a dashboard with alerts before launch.
Secrets out of the repo
Device and provider keys live in the cloud's secret store, never in the app bundle.
Route optimization and carrier integrations would be new work. Everything else on this page has run in production.
Need operations to see the same thing at once?
Tell me what dispatch uses today and where the updates get lost. I will come back with what I would make real-time first.