Shared inbox · AI triageAn AI-triaged operations inbox for a customs clearance team
The problem
A small clearance team's shared inbox received everything — supplier invoices, carrier notices, client queries, quote requests, statements — and every message needed a person to read it just to decide whether it mattered. Team addresses accumulated across a brand transition, multiplying the places mail could land and quietly go stale.
What we built
All legacy team addresses route to one monitored operations inbox. An AI layer reads each inbound message and classifies it: routine correspondence is filed and summarised; quote requests and genuine escalations are surfaced to a person immediately. Nothing is auto-replied without human review, and nothing silently disappears — a catch-all guarantees every legacy address still lands somewhere monitored.
What changed
The team stopped monitoring half a dozen mailboxes. Quote requests get seen the day they arrive instead of whenever someone next checked an old address. The clearance operators spend their inbox time on the messages that need judgment, not on sorting.
Honest note: getting here wasn't one clean build. The brand transition surfaced dead addresses that were bouncing external mail before the catch-all closed the gap — finding and fixing that class of quiet failure is a real part of what this work looks like.
Multi-tenant platform · BookingsA freight operations platform our own staff use daily
The problem
Container cartage bookings, client records and job files lived across spreadsheets, email threads and individual memory. Adding a new operator meant weeks of "ask the person who knows." Serving a second brand or client group meant duplicating everything.
What we built
A multi-tenant operations platform — bookings, clients, jobs and staff access — where each business unit is a tenant with row-level data isolation, not a separate copy of the system. New staff get scoped logins in minutes. A container cartage module handles quote-to-booking flow for the cartage business, priced against real market rate data.
What changed
Onboarding a new client group is now a provisioning task, not a rebuild. Operators in different cities work the same job queue with the same data. When a booking question comes up, the answer is in the system rather than in someone's inbox.
Honest note: the platform is ours and purpose-built for freight operations. We'd scope your version against your processes — the point of this case study is that we run this daily, not that one size fits all.
Email infrastructure · MigrationA self-hosted, multi-domain email cutover under live operational load
The problem
Three trading brands needed email moved off a patchwork of third-party providers onto infrastructure we control — without dropping creditor correspondence, client quotes or compliance mail mid-transition, and without staff losing access for days.
What we built
A self-hosted mail stack serving multiple domains, with per-person mailboxes, alias mapping from every legacy address, DKIM/SPF/DMARC alignment verified at major receivers, and webmail for staff who work in a browser. Cutover was staged: receive first, verify, then switch sending, then retire the old provider — with rollback available at each step.
What changed
Staff across three brands work from one mail system with their own addresses. Legacy addresses keep working through forwarding instead of bouncing. The operating cost of third-party mailbox licences went away.
Honest note: real cutovers surface real faults. Ours included a webmail database error a team member caught the morning after go-live and a password that hadn't carried across for one operator — both fixed same-day. We tell you this because "zero-issue migration" is a claim you should distrust.
Want this level of candour about your own operation?
The audit is where we start: one process mapped, a costed automation plan with acceptance checks, yours to keep whoever builds it. Fixed price, no retainer, no lock-in.
Book an automation audit — AUD $495