← ALL CASE STUDIESCLOUD & DEVOPS · 2025

Freight platform rebuild.

How a national freight operator went from a failing monolith to event-driven services — without stopping a single truck.

CLIENTNational freight operator
SERVICESCloud, custom software, SRE
TIMELINE14 months, team of 7
HEADLINE RESULTDispatch latency down 94%
RJP-2025-014 · DISPATCH CONTROLALL SYSTEMS NOMINAL
−94%DISPATCH LATENCY · P95 · ROLLING 30 DAYS
DISPATCH ▸ OKROUTING ▸ OKBILLING ▸ OKTELEMETRY ▸ OK

01The challenge

A 15-year-old monolith scheduled every truck in the fleet. Deploys took a weekend, a single bad query could stall dispatch nationwide, and the vendor who built it no longer existed. Peak-season outages were costing seven figures per incident.

02The approach

We strangled the monolith service by service: dispatch, routing, billing, and telemetry became independent event-driven services on AWS, each cut over behind a feature flag with the old system as fallback. No big-bang migration, no downtime windows — the fleet never noticed.

BEFORE
MONOLITH
15 YRS · WEEKEND DEPLOYS
→
AFTER — EVENT-DRIVEN SERVICES
DISPATCH
ROUTING
BILLING
TELEMETRY

03The result

94%LOWER DISPATCH LATENCY
40×MORE FREQUENT DEPLOYS
0PEAK-SEASON OUTAGES SINCE

Two years on, our SRE team still runs the platform under a managed SLA. The client's own engineers ship to it daily — the codebase was theirs from day one.

"They run like part of our own team. Five years in, Rajrappa still owns the systems our entire operation depends on."
CT
CTO, CLIENT NAME
NEXT CASE STUDYBank onboarding, automated →
Start a project