eeztech
Last-mile deliveryGoal: launch a product5 months to launch

A same-day delivery platform — routing, batching and couriers in one system

A same-day delivery company needed its own platform: apps for couriers and riders, automatic routing and order batching, fast onboarding for new couriers, and a control room that shows every delivery live.

Client
Same-day last-mile delivery company
Their team
Founders, head of operations, dispatch team of 6
Our team
Product manager, 6 engineers, designer, QA
Timeline
5 months to launch · continuous development for 2+ years
3 h 40 → 1 h 55
average time from order to doorstep
−27%
cost per delivered order
3 days → 4 h
to get a new courier on the road
1,400
couriers and riders on the platform

The business goal

Deliver the same day at a cost per order that makes the business work: fewer kilometres per order, more orders per courier, and new couriers on the road within hours, not days.

Agreed in the outcome brief
Criteria
Order-to-doorstep time, cost per delivered order, orders per courier per shift
Baseline
3 h 40 min; 11 orders per courier shift — 4 weeks of manual dispatch
Target
Under 2 h; 18+ orders per shift
Pilot exit
Routing and batching beat manual dispatch in one city for 3 weeks in a row

Where it was

Orders were assigned by hand from a shared screen, couriers called the office for addresses, and onboarding a new courier meant paperwork and a day of training. Nobody could see in real time where a delivery was.

Orders
from shops and marketplaces
Manual dispatch
calls, chats, spreadsheets
Couriers
one order at a time

What we did

  1. 1Discovery · 3 weeksRode along with couriers and sat with dispatchers; defined the delivery flow, zones and what “late” means.
  2. 2First version · 5 monthsCourier and rider apps, an order API for partner shops, automatic assignment to the nearest free courier, a live map for operations.
  3. 3Routing & batching · 2 monthsOrders going in the same direction grouped into one trip with an optimised route; time windows and vehicle types taken into account.
  4. 4Onboarding & monitoring · ongoingSelf-service courier onboarding with document checks and in-app training; monitoring of courier performance and delivery quality.

What did not work

Automatic batching at the evening peak: on the map the batches looked optimal, but couriers lost time at building entrances and parking. We added real hand-over times per address to the model and let dispatchers split a batch with one tap.

Who did what

The client’s team
Defined delivery zones and service levels, ran the courier fleet, brought feedback from the street every week.
Our team
Product, design, apps, routing and batching, the operations dashboard, support.
Together
A weekly review of delivery time, cost per order and courier feedback.

Before and after

NumberBeforeAfter
Order to doorstep3 h 40 min1 h 55 min
Orders per courier per shift1119
Orders delivered in batches0%62%
Time to onboard a courier3 days4 h
Cost per delivered orderbaseline−27%

Under the hood

Courier and rider mobile apps; partner order API; routing and batching service (vehicle routing with time windows and real hand-over times); real-time location tracking; operations dashboard with a live map; onboarding with document checks; analytics on courier performance and delivery quality.
We stopped running delivery from a group chat. The platform tells couriers where to go — and tells us when something is about to go wrong.
Head of Operations, same-day delivery company · name withheld at the client’s request
Similar cases
Field service9 days → same dayAn internal app the ops team built with no-code and AI tools — made solid enough for 300 techniciansRetail−31%Store replenishment suggestions instead of spreadsheets and gut feelingCleaning robots86% → 95%One fleet console for rented cleaning robots — problems found before the customer calls
A similar situation?
Get a starting plan for your situation in three questions — or talk it through with us.