A transport management platform built from scratch, for six different jobs
National carrier. From discovery to release: web platform, warehouse app, driver app, route optimisation with predictive AI.
- My roleSenior Digital Product Owner — discovery, requirements, backlog, team coordination, release
- DurationAbout two and a half years
- AreasDiscovery · Solution design · Predictive AI · Robotic warehouse integration · Mobile apps
A national logistics operator was handling its e-commerce clients’ orders and its delivery route planning with disconnected tools. There was no single system: each department worked on its own piece, and information moved between departments by hand.
The initial ask was “we need a TMS”. The real problem, which surfaced in the interviews, was different: six distinct jobs — warehouse operator, driver, customer support, finance, logistics, operations — with needs, priorities and working conditions that were incompatible with one another. The warehouse operator works standing up, wearing gloves, holding a barcode scanner. The driver is on the road, often with one hand free. Finance needs clean data at month end. A single tool designed for the average of all of them would have been unusable for each of them.
- 01
Explore and understand
Interviews with every user group and every department head, on-site observation, as-is process mapping. This is where it became clear that the project was not one system but six experiences on a single data backbone.
- 02
Design
Functional architecture and solution design: a web platform with role-based sections and permissions — the warehouse operator does not see invoicing, finance does not see parcel handling — plus two mobile applications designed around real working conditions.
- 03
Build
Prototypes validated with end users before development, department by department. The heaviest corrections came at this stage, when changing course was still cheap.
- 04
Deliver
Backlog and sprints over roughly two and a half years, coordinating the technical teams and keeping stakeholders with conflicting priorities aligned. Progressive release, department by department, rather than a single cutover.
- Web platform orchestrating every order coming from e-commerce clients, with role-based views and permissions.
- Warehouse app for parcels in transit: barcode scanning and automatic assignment to the most efficient route or trip.
- Last-mile driver app: delivery outcomes, signature capture, PIN handling.
- Predictive AI for dynamic load forecasting and optimisation of forwarding and delivery routes.
- Integrations with the robotic warehouse, with clients’ e-commerce platforms, with shipment progress notifications, and with accounts payable and receivable.
The pattern recurs whenever an operational process spans very different jobs. The temptation is to build one uniform system; what works is a single data backbone with separate experiences. Recognising this during discovery, rather than after the first release, is the difference between a project that finishes and one that stalls.
The client is not named, for confidentiality. Sector and scale are real, as is everything else here.
Recognise the situation?
If something here resembles your case, send me a couple of lines. If it is not my ground, I will say so straight away.